Vous êtes sur la page 1sur 36

Agency Theory: An Assessment and Review

Author(s): Kathleen M. Eisenhardt


Source: The Academy of Management Review, Vol. 14, No. 1 (Jan., 1989), pp. 57-74
Published by: Academy of Management
Stable URL: http://www.jstor.org/stable/258191
Accessed: 07-09-2015 13:38 UTC
REFERENCES
Linked references are available on JSTOR for this article: http://www.jstor.org/stable/258191?
seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

YouruseoftheJSTORarchiveindicatesyouracceptanceoftheTerms&ConditionsofUse,availableat
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTORisanotforprofitservicethathelpsscholars,researchers,andstudentsdiscover,use,andbuilduponawiderangeof
contentinatrusteddigitalarchive.Weuseinformationtechnologyandtoolstoincreaseproductivityandfacilitatenewformsof
scholarship.FormoreinformationaboutJSTOR,pleasecontactsupport@jstor.org.

Academy of Management iscollaboratingwithJSTORtodigitize,preserveandextendaccessto The Academy of Management


Review.

http://www.jstor.org

ontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

?Academy of Management Review, 1989,Vol. 14,No. 1, 5774.

Agency

Theory:
and

An Assessment
Review

KATHLEENM.EISENHARDT
StanfordUniversity
Agencytheoryisanimportant,yetcontroversial,theory.Thispaper
reviewsagencytheory,itscontributionstoorganizationtheory,andthe
extantempiricalworkanddevelopstestablepropositions.Theconclusions
arethatagencytheory(a)offersuniqueinsightintoinformationsystems,
outcomeuncertainty,incentives,andriskand(b)isanempiricallyvalid
perspective,particularlywhencoupledwithcomplementaryperspectives.
Theprincipalrecommendationistoincorporateanagencyperspectivein
studiesofthemanyproblemshavingacooperativestructure.
OnedayDengXiaopingdecidedtotake
his grandson to visit Mao. "Call me
granduncle," Mao offered warmly.
"Oh, I certainly couldn't do that,
ChairmanMao,"theawestruckchild
replied. "Why don't you give him an
apple?"suggestedDeng.Nosoonerhad
Mao done so than the boy happily
chirped,"Ohthankyou,Granduncle."
"You see," said Deng, "what in
centives

can

achieve."
("Capitalism,"1984,p.62)

purposes of this paper are to describe


agencytheoryandtoindicatewaysinwhich
organizational researchers can use its
insights. The paper is organized around
four questions that are germane to
organizational research. The first asks the
deceptivelysimplequestion,Whatisagency
theory? Often, the technical style, math
ematics, and tautological reasoning of the
agency literature can obscure the theory.
More

over, the agency literature is split into two


place"(Jensen,1983,p.324).Itsdetractors
Agencytheoryhasbeenusedbyscholars
call it trivial, dehumanizing, and even
in accounting (e.g., Demski & Feltham,
1978),eco
nomics
nance
Lal,
ence
ior(e.g.,
and
Yet,itisstillsurroundedbycontroversy.
ponentsarguethatarevolutionisathand
and that "the foundation for a powerful
theoryoforganizationsisbeingputinto

"dangerous"(Perrow,1986,p.235).
Which is it:grand theory orgreat
sham?
The

camps(Jensen,1983),leadingto

agencytheorycontributetoorganizational

differencesininterpretation.For

theory?ProponentssuchasRoss(1973,

example,BarneyandOuchi(1986)

p.134)arguedthat"examplesofagency

arguedthatagencytheoryemphasizes

areuniversal."Yetotherscholarssuch

howcapitalmarketscanaffectthe

asPerrow(1986)claimedthatagency

firm,whereasotherauthorsmadeno

theoryaddressesnoclearproblems,and

referencetocapitalmarketsatall

HirschandFriedman(1986)calledit

(Anderson,1985;Demski&Feltham,

excessivelynarrow,focusingonlyon

1978;Eccles,1985;Eisenhardt,1985).

stockprice.Foreconomists,long

Thesecondquestionis,Whatdoes

accustomedtotreatingtheor

57

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

ganization
firm,agency
fororganizational
theoryisnotsoobvious.
Thethirdquestion
icallyvalid?Thepoweroftheempirical
onagency
nomena
light
"hardlysubject
triestoexplain
224).Perrow
beingunrealisticallyonesided
neglect
Thefinalquestion
are
use
agency
requires
theagency
leverage.
Theprincipal
present
tionsofthetheoryto
andevaluate
overallconclusion
fuladdition
agency
tainty,
novel
and

theory,
mentary

Origins ofAgency Theory


During the 1960s and early 1970s,
economistsexploredrisksharingamong
individualsorgroups(e.g.,Arrow,1971;
Wilson,1968).Thisliteraturedescribed
the risksharing problem as one that
arises when cooperating parties have
different attitudes toward risk. Agency
theory broadened this risksharing
literaturetoincludethesocalledagency
problem that occurs when cooperating
partieshavedifferentgoalsanddi

58

vision
1973).Specifically,
theubiquitous
oneparty(theprincipal)delegates
other
Agency
shipusing
Meckling,
Agency
twoproblemsthatcanoccurinagency
ships.Thefirstistheagency
when
agent
fortheprincipaltoverifywhat
tuallydoing.Theproblem
cipalcannotverifythattheagent
appropriately.
sharing
agenthavedifferentattitudestoward
problem
maypreferdifferentactionsbecause
ferentriskpreferences.
Because
governing

paland
determining the most efficient contract
governing

the

principalagent

relationship given assumptions about


people(e.g.,selfinterest,
bounded
tions(e.g.,
information
which
question
tract
moreefficientthananoutcomeoriented
tract(e.g.,
ofpropertyrights,marketgovernance)?
view
Theagencystructureisapplicable
ofsettings,rangingfrommacrolevel
asregulatory
nomena such as blame, impression
management,lying,andotherexpressions
of selfinterest. Most frequently, agency
theoryhasbeenappliedtoorganizational
phenomena

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

Table 1
Agency TheoryOverview
Key idea
should

Principalagent relationships
reflect efficient
organization
of
information and risk
bearing
costs

Unitof
analysis
Human
assumptions
Organizational
assumptions

Information

Contract between principal


and
agent

positivistandprincipalagent(Jensen,
1983).Thetwostreamsshareacommon
unitofanalysis:thecontractbetweenthe
principalandtheagent.Theyalsoshare
commonassumptionsaboutpeople,
organizations,andinformation.However,
theydifferintheirmathematicalrigor,
dependentvariable,andstyle.

assumption
Contracting
problems
Problem
domain

such
1988;Eisenhardt,
fication
boardrelationships(e.g.,
Kosnik,
tures(e.g.,Argawal&Mandelker,
&Meckling,
son,
1988;Zenger,
agency
basic
agent
ior,but
tudes

as

1985;Eccles,

Fromitsrootsininformationeconomics,agencytheoryhasdevelopedalongtwo
lines:

59

PositivistAgency Theory
Positivistresearchershavefocusedonidentifyingsituationsinwhichtheprincipaland
agent
arelikelytohave
scribingthegovernance
the
search
agent
havefocusedalmost
case
owners
tions(Berle&Means,
Threearticleshavebeenparticularlyinfluen
tial.Jensen
ownership
inghowequity
managers'

agent's

oftheprincipalagent

is
research

and

(1980)discussed
labormarkets
are
top
scribed
informationsystemthatthestockholders
largecorporations
portunismoftopexecutives.
leagues
extended
suchasgolden
ing.
Froma
stream
ing
agency
thisinterestas"whycertain
tionsarise."
ernance
positiviststream.

used
executives.

thegovernance

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

comebasedcontractsareeffectiveincurbingagentopportunism.Theargumentisthat
suchcontractscoalignthepreferencesofagentswiththoseoftheprincipalbecausethe
rewardsforbothdependonthesameactions,and,therefore,theconflictsofself
interestbetweenprincipalandagentarereduced.Forexample,JensenandMeckling
(1976)describedhowincreasingthefirmownershipofthemanagersdecreases
managerialopportunism.Informalterms,
Proposition1:Whenthecontractbetweentheprincipalandagentisoutcomebased,the
agentismorelikelytobehaveintheinterestsoftheprincipal.

Thesecondpropositionisthatinformationsystemsalsocurbagentopportunism.
Theargumenthereisthat,sinceinformationsystemsinformtheprincipalaboutwhat
theagentisactuallydoing,theyarelikelytocurbagentopportunismbecausethe
agentwillrealizethatheorshecannotdeceivetheprincipal.Forexample,Fama(1980)
describedtheinformationeffectsofefficientcapitalandlabormarketsonmanagerial
opportunism,andFamaandJensen(1983)describedtheinformationrolethatboards
ofdirectorsplayincontrollingmanagerialbehavior.Informalterms,

Proposition2:Whentheprincipal
tiontoverifyagent
likelytobehave

therole

has

problem

mecha

Atitsbest,positivistagency
gardedas
more
1983).However,
nizational
Michaels,
bymicroeconomists
rigor
agency
search
terest ("MeetMike," 1988).

complex

(Jensen,
theory

60

PrincipalAgentResearch
Principalagentresearchersareconcernedwithageneraltheoryoftheprincipalagent
re
lationship,atheorythatcanbe
ployeremployee,lawyerclient,
andotheragencyrelationships
1978).Characteristicofformaltheory,theprincipalagentparadigminvolvescareful
specifi
cation
logical
Incomparison
cipalagent
and,
scholars.
theory
focused
widely
palagent
greater
tions.Incontrast,
cused
theowner/CEO
ration.
cludesmanymoretestableimplications.
Fororganizational
provide
ofthe
Rather,
streams
identifiesvariouscontractalternatives,
cipalagent
themostefficientundervarying
come

ofassumpti
deduction

(Perrow,

knownpos
interest

Finally,

backgrou
theory.How

andothervariables
Thefocusoftheprincipalagent
ondetermining
versus
agent.
between
suredoutcome,andanagentwho
averse
behind
whoare
should

outcome,
Thesimple

thanthepri
amorerisk
be

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

capable
shouldberiskneutral.)Theapproach
plemodel
(e.g.,
simplecaseofcomplete
principal
Given
behavior,
haviorismostefficient.Anoutcomebased
tractwouldneedlessly
whoisassumedtobemoreriskaverse
principal.
Thesecond
notknow
Given
mayormaynothave
agencyproblemarisesbecause
andthe
principal
haved
twoaspects
Moralhazard
theagent.
maysimplynotputforththeagreedupon
That
moral hazard occurs when a research scientist
works
pany
corporatemanagement
scientist
fersto
agent.Theargument
claim
orsheishired.Adverse

Demski&Feltham,

thattheprincipal

the

appropriately.

is,
onapersonal
time,buttheresearch

isa
th
tohave

theprincipal
skills
while
verseselectionoccurswhenaresearchscientistclaimstohaveexperienceinascientific

orabilities
theagent

specialtyandtheemployercannotjudgewhetherthisisthecase.
Inthecaseofunobservablebehavior(duetomoralhcazardoradverseselection),the
principal has two options. One is to discover the agent's behavior by investing in
informationsystems
such as budgeting systems, reporting procedures, boards of directors, and additional
layersofmanagement.Suchinvestmentsrevealtheagent'sbehaviortotheprincipal,and
thesituationrevertstothecompleteinformationcase.Informalterms,
Proposition
tivelyrelatedtobehaviorbased
negatively

Theotheroptionistocontractontheoutcomes
oftheagent's
contractmotivatesbehavior
theagent's
pal,butat
agent.Theissue
areonlypartlyafunctionofbehaviors.Governmentpolicies,economicclimate,
competitorac
tions,
cause
The
resulting
not
onlythe
thatmustbebornebysomeone.Whenoutcomeuncertaintyislow,thecostsofshifting
risktotheagentarelowandoutcomebasedcontractsareattractive.However,as
uncertaintyincreases,itbecomesincreasinglyexpensivetoshiftriskdespitethe
motivationalbenefitsofoutcomebasedcontracts.Informalterms,
Proposition4: Outcomeuncertainty ispositively relatedto behaviorbasedcontracts and
negativelyrelatedtooutcomebasedcontracts.

Thissimpleagency
invarying
&Feltham,
strom,1979;Shavell,
ofprincipalagent
tween(a)thecostofmeasuring
thecostofmeasuring
risktotheagent.
Anumberofextensions
arepossible.
riskaverse
Research
catesthat

technologicalchange,and
uncontrollable

61

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

attitudes.
lessriskaverse(e.g.,
comesmoreattractivetopassrisktotheagent
usinganoutcomebased
astheagentbecomes
creasingly
formal terms,
Proposition 5: The risk aversion of the
agent is positively related to behavior
basedcontractsandnegativelyrelatedto
outcomebasedcontracts.

Similarly,astheprincipalbecomesmore
riskaverse,itisincreasinglyattractiveto
passrisktotheagent.Informalterms,
Proposition6:Theriskaversion
isnegatively
tracts
based

Another
ofgoalconflictbetween
(e.g.,
highly
1979)orinsituations
waytoselfless
isnogoal
principal
orherbehavior
decreases,
imperative
theissue
Under
behaviorbased
tive.Informalterms,
Proposition7:Thegoalconflictbetween
principalandagentisnegativelyrelated
to behaviorbased contracts and
positively related to outcomebased
contracts.

Anothersetofextensionsrelatestothe
task performed by the agent. For
example,theprogammabilityofthetask
islikelytoinfluencetheease
ofmeasuring
Programmability

whichappropriatebehaviorbythe
bespecified

aretailsalescashier
thanthatofahightechnology
Theargumentisthatthebehavior
gaged
serve
grammedthetask,themoreattractiveare

longtimetocomplete,involvejointorteam
effort,orproducesoftoutcomes.Inthese
circumstances, outcomes are either
difficulttomeasureordifficulttomeasure
withinapracticalamountoftime.When

behaviorbased contracts because

outcomes are measured with difficulty,

information abouttheagent'sbehavioris

outcomebased contracts are less

more readily determined. Very

attractive.Incontrast,whenoutcomesare

programmed tasks readily reveal agent

readilymeasured,outcomebasedcontracts

behavior, and the situation revertsto the

aremoreattractive.Informalterms,

complete information case. Thus, retail


salesclerksaremorelikelytobepaidvia
behaviorbased contracting (e.g., hourly
wages), whereas entrepreneurs are more
likelytobecompen
satedwithoutcomebasedcontracts
ownership).Informalterms,
Proposition
tivelyrelatedtobehaviorbasedcontracts
negatively

Another task characteristic is the


measurability of the outcome (Anderson,
1985;Eisenhardt,1985).Thesimplemodel
assumes that outcomes are easily
measured.However,sometasksrequirea

Proposition9:Outcomemeasurabilityis
negatively related to behaviorbased
contracts and positively related to
outcomebasedcontracts.

Finally, it seems reasonable that when


principals and agents engage in a long
term relationship, it is likely that the
principal will learn about theagent(e.g.,
Lambert, 1983) and so will be able to
assessbehaviormorereadily.Conversely,
in shortterm agency relationships, the
informationasymmetrybetweenprincipal
and agent is likely to be greater, thus
makingout

62

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

comebasedcontractsmoreattractive.Informalterms,
Proposition10:Thelengthoftheagencyrelationshipispositivelyrelatedtobehaviorbased
contractsandnegativelyrelatedtooutcomebasedcontracts.

Agency Theoryand the


Organizational Literature
DespitePerrow's(1986)assertionthatagencytheoryisverydifferentfromorganization
theory,agencytheoryhasseverallinkstomainstreamorganizationperspectives(seeTable

2).Atitsroots,agencytheoryisconsistentwiththeclassicworksofBarnard(1938)onthe
natureofco
operative
ontheinducements
ployment
theheart
inherent
ences
sential
Agency
elsoforganizations.
perspectives
atthe
organizational
1981).Also,
asymmetry
participants
enceisthatinpoliticalmodels
resolved
coalitionsthe
science.
through
pricemechanism
Agency
tionprocessing
ory(Chandler,
&Lorsch,
tiontheories.Theyassume
boundedly
tributedasymmetrically
zation.Theyalsoareefficiency
theyuseefficientprocessing
criterionforchoosing
forms(Galbraith,
thetwoistheirfocus:Incontingency
searchers
turingofreporting
making responsibilities (e.g., Galbraith,

engage

individual

1973;

Lawrence
theorytheyareconcerned
structuring
fromthese
terns.Forexample,
wewould
organized

Table2
ComparisonofAgency TheoryAssumptionsandOrganizationalPerspectives

Assumption
Selfinterest
Goalconflict
Boundedrationality
Informationasymmetry
Preeminenceofefficiency
Riskaversion
Informationasacommodity

63

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

Usingagency
withwhether
ture
tives.
controlliterature(e.g.,
Forexample,
chi's
tionships
susoutcome
theory'slinkingtaskprogrammability
surability
hardt,
tionships
iorcontrol,and
outcomes)
Ouchi's
frameworktoinclude
assuming
agency
gruencebetweenpeopleand,therefore,
duced
Motivationissuesdisappear.
ences
zationalcontrolliteraturearethe
tionsofprincipalandagent
outcomeuncertainty(Propositions
ties
(Williamson,
chi(1986),thetheories

are

(1979)linking

1985).Tha

(1979)ex

theory.C
need

with

interest
have
similar
archiesroughlycorrespondtobehaviorbasedcontracts,andmarketscorrespondto
outcomebasedcontracts.However,thetwotheoriesarisefromdifferenttraditionsin
economics(Spence,1975):Intransactioncosttheorizingweareconcernedwith
organizationalboundaries,whereasinagencytheorizingthecontractbetween
cooperatingparties,regardlessofboundary,ishighlighted.However,themostimpor
tantdifferenceisthateachtheoryincludesuniqueindependentvariables.In
transactioncosttheorytheseareassetspecificityandsmall

64
numbersbargaining.Inagencytheorytherearetheriskattitudesoftheprincipaland
agent, outcome uncertainty, and information systems. Thus, the two theories share a
parentage in economics, but each has its own focus and several unique independent
variables.

ContributionsofAgency Theory
Agencytheory
ofincentives
thinking(Perrow,
usthatmuch
likeitornot,
theoryalso
commonproblemstructureacrossresearchtopics.AsBarneyandOuchi(1986)described
it,organizationresearchhasbecomeincreasingly
topic,ratherthantheory,centered.
oryremindsusthatcommonproblem
doexistacross
sultsfromone
gration)maybegermane
monproblem
Agency
butionstoorganizational
treatment
formationisregarded
cost,
important
such
tors,
supervision,which
research.
caninvest
controlagent
Anillustrationofthisisexecutive
tion.Anumber
expressed

and
asbudgeting,
and

based
Stevenson,
1984).However,
isnot
should
including

executive

surprising
be

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

richerinformationsystemscontrol
managerialopportunismand,therefore,
leadtolessperfor
mancecontingent
Oneparticularlyrelevantinformationsystem
formonitoringexecutivebehaviorsistheboard
ofdirectors.Fromanagency
canbe
holder
boards provide richer information,
compensa
tion
mance.
ecutives
onknowledge
likely.Executiveswouldthenberewarded
taking
risk/highpotential
beunsuccessful.
richer
likelytoengage
withstockholders'interests.Forexample,
anagencyviewpoint,behaviorssuch
greenmail
tobenefitthemanager
ers,arelesslikelywhen
itors
therichnessofboard
suredinterms
quency
subcommittees,
long
managerial
ber
ownership
Asecond
riskimplications.
have

prosperity,
outcome, and thatfutureisonly partly
controlled
byorganization
fects
gence
vation
tends
ramifications
implications

65

viewed
intermsofinability
isthatoutcomeuncertainty
ences
ence
Verticalintegrationprovidesanillustration.
Forexample,
thattechnological
notaffectthe"makeorbuy"decision
nentsinalargeautomobile
cipal
explain
framework.However,
withagencythinkingifthemanagers
tomobile
sumption
relative
nent).According
predictthatsuchariskneutralprincipal
tively
whichwasWalkerandWeber'sresult.

thiscase,
limited
uncertainty:
large.Inthiscase,
mayberiskaverse
agency
agers
tainty.Inparticular,
morelikelytochoose
transferringrisktothesupplying
agency
ersare
haviorbased
ecutives
based

EmpiricalResults
Researchers in several disciplines have
undertaken empirical studies of agency
theory. These studies, mirroring the two
streams of theoretical agency research,

reverse

areinTable3.

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

93
to

;5
to

0
0

U)

to

tj

0
ti

tm64

>

0
u

5
U)
(1)

>

(1)

U)
(1)
0

to

CV)

66

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52UTCAllusesubjecttoJSTORTermsandConditions

ui

a)

U2.1
0
tj

ti
U2

a)
0)

ts0
0
6

U2

a) k4
0
(
U2

CD
CD

a
)

U2

00

u
P
C

04

0
U
2

tj

ti

>
U2
0

.
1

ti
co

:,:

I.

U2

a)

ui
0

U2
U2

a)
a)
U2

ui

a)
>
t))

U2

a)> a)
a)

_C:
0

0
U2

a)
a)
0a)d
>

rn
U2

U2

U2

a)

U2

I
.1

a) a)
U2 J,
U2

>1

67

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions

Results ofthePositivist Stream


Inthepositiviststream,thecommon
approachistoidentifyapolicyor
behaviorinwhichstockholderand
managementinterestsdivergeandthento
demonstratethatinformationsystemsor
outcomebasedincentivessolvetheagency
problem.Thatis,thesemechanisms
coalignmanagerialbehaviorswithowner
preferences.Consistentwiththepositivist
tradition,mostofthesestudiesconcern
theseparationofownershipfrom
managementinlargecorporations,and
theyusesecondarysourcedatathatare
availableforlargefirms.
Oneoftheearlieststudiesofthistype
wasconductedbyAmihudandLev
(1981).Theseresearchersexploredwhy
firmsengageinconglomeratemergers.
Ingeneral,conglomeratemergersarenot
intheinterestsofthestockholders
because,typically,stockholderscandiver
sifydirectlythroughtheirstockportfolio.
Incon
trast,conglomeratemergersmay
tomanagers
todiversify
mergersareanarenainwhichownerand
manager interests diverge. Specifically,
these authors linked merger and
diversificationbehaviorstowhetherthe
firm was owner controlled (i.e., had a
major stockholder) or manager
controlled (i.e., had no major
stockholder).Consistentwith

agencytheoryarguments
1976),managercontrolled
nificantly
lated)acquisitions
Alongthesamelines,WalkingandLong
(1984)studiedmanagers'
bids.Theirsample
porations
between
takeover
butitmaybeintheinterestsofmanagers
becausetheycanlosetheirjobsduringa
takeover.Consistentwithagencytheory
(Jensen&Meckling,1976),theauthors
foundthatmanagerswhohavesubstantial
equitypositionswithin

68
theirfirms(outcomebased
likelytoresisttakeoverbids.
Theeffectsofmarketdiscipline
lationships
study
(principals)
andgas
taxand
toassess
nance
tensive
incentives
withagency
foundthatlongrunreputationeffectsofthe
marketcoalignedtheshortrunbehaviorsof
the general partner with the limited
partners'welfare.
Kosnik(1987)examined
mechanism
boardofdirectors.Kosnikstudied
corporations
tween
agency
ticsto
(payinggreenmailisconsidered
holders'interests).Aspredicted
ory(Fama&Jensen,
thatresisted
ofoutsidedirectorsandahigher

outside
Ina
(1987)examined
firmsecurities
tweenstockholdersandmanagement.
Specifically,theystudiedtherelationship
betweenstockandstockoptionholdingsof
executivesandwhetheracquisitionand
financingdecisionsweremadeconsistent
withtheinterestsofstockholders.Ingeneral,
managerspreferlower
riskacquisitions
Argawal
sample
acquisitions
1982.Consistent
&Meckling,
(outcomebased

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions

sitionandfinancingdecisionsthatwere
moreconsistentwithstockholderinterest.
Thatis,executivestockholdings
appearedtocoalignmanagerial
preferenceswiththoseofstockholders.
Singh and Harianto (in press) studied
goldenparachutesinamatchedsampleof
84 Fortune 500 firms. Their study
includedvariablesfrombothagencyand
managerialistperspectives.
Consistentwithagencytheory(Jensen&
Meckling,1976;Fama&Jensen,1983),
theauthorsfoundthatgoldenparachutes
are used to coalign executive interests
with those of stockholders in takeover
situations, and they are seen as an al
ternative outcomebased contract to
executive stock ownership. Specifically,
theauthorsfoundthatgoldenparachutes
werepositivelyassociatedwithahigher

probability of a takeover attempt and


negativelyassociatedwithexecutivestock
holdings.
Finally,Barney(1988)explored
whetheremployeestockownership
reducesafirm'scostofequitycapital.
Consistentwithagencytheory(Jensen&
Meckling,1976),Barneyarguedthat
employeestockownership(outcome
basedcontract)wouldcoalignthe
interestsofemployeeswithstockholders.
Usingefficientcapitalmarket
assumptions,hefurtherarguedthatthis
coalignmentwouldbereflectedinthe
marketthrougha
lowercostofequity.Although
directlytest
areconsistent
Insummary,
ofagency
topexecutives
interests
debtversus
divestitures,
problems
suchasgolden
inpress)andexecutive
&Mandelker,
(b)through
(Kosnik, 1987) and efficient markets
(Barney, 1988; Wolfson, 1985). Overall,
thesestudiessup

69
port the positivist propositions described
earlier. Similarly, laboratory studies by
Dejongandcolleagues(1985),whicharenot
reviewedhere,arealsosupportive.

ResultsofthePrincipalAgentStream
The principalagent stream is more
directlyfocusedonthecontractbetweenthe
principal and the agent. Whereas the
positiviststreamlaysthefoundation(thatis,

thatagencyproblemsexistandthatvarious
contractalternativesareavail
able),theprincipalagentstream
mostefficientcontractalternativeinagiven
uation.Thecommon
istouseasubset
taskprogrammability,
outcomeuncertainty
tractisbehavior
lyingassumption
willchoosethemostefficientcontract,
althoughefficiencyisnotdirectlytested.
Inonestudy,Anderson(1985)probed
verticalintegrationusingatransactioncost
perspective
withagency
ined
resentative
sales
electronics
toryvariable
culty
amountofnonselling
Consistent
able
sales
Inotherstudies,
inedthechoice
based)andsalary(behaviorbased)
tionofsalespeople
study
while
agency
dictions.
predictions
tionsystems
andoutcome

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions

bynumberof

thesalaryversus

competitorsand

commission

failurerates)sig

choice.
aswell.
Conlon
tended

nificantlypredict

ting.Theyusedamultiperioddesign
theorymakes
agency
contributionsto
withagency
organization
foundthatinformationsystems
whether
theory,istestable,
agent's
and
performancecontingent(outcomebased)
Theyalso
70
dictions.
Finally,
develop
pricing.
13large
workbased
prescribe the
conditions
under
which
various
sourcing
bothefficient
framework
(arguably
andthe
contract)
transferpricing
Insummary,
agent
informationsystems
cles,1985;Eisenhardt,
tainty
ability
time(Conlon
grammability
Moreover,
variety
secondary
and interviews.

Recommenda
tions
forAgency
Theor
y
Resear
ch
Asargued
above,agency

hasempiricalsupport. newconcepts
Overall, it seems as
cal
reasonable to urge
ances,
the adoption of an
Studying
agency

theory ticularlyopportune
perspective when inmeasuring
investigatingthemany worksofKahneman
CrimmonandWehrung(1986),andMarchand
problemsthathavea
Shapira
principalagent
canmeasure
structure.Fivespecific realistically.
recommendations are measures
indirectmeasures
outlined below for
teristics
usingagencytheoryin characteristics
organizational re MarchandShapira,
search.

Focuson
Information
Systems,

KeyonTheory
RelevantContexts
Organizational

theory usually is
explored in settings

Outcome

in which the theory

Uncertainty,

appears to have

andRisk

greatest relevance.

For

example,
McGrath,
institutional and
thatresearch
resource dependence
Using
theories were devel
agency
should
oped primarily in
uncertainty,
large,

public
makethemostuniquecontributiontoorganiza
bureaucracies in
tionalresearch,
whichefficiencymay
piricalattention
not have been a
searchersplace
ordertoadvance
pressing

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions

concern.
thesame
theoryrelevant
Agency
whichcontracting

include
tialgoal
suchthatagent
ers
als,

im

uncertainty
theory
and
tries);and
jobsinwhich
Byemphasizing
useagency
leverage
tested.Topicssuchasinnovationand
suchastechnologybased
attractive
between
jobs
cult.

71

ExpandtoRicherContexts
Perrow(1986)andothershavecriticizedagency
theoryforbeing
few testable
implications.
Although
these
criti

cisms
search
Thus,
richerandmorecomplexrange
Twoareas
istoapply
tional
asymmetry
tions.Examples
management
andother
blame
theory might
contribute

an

overall framework
in which to place
thesevariousforms
of

selfinterest,

leading to a better
understanding of
when

such

behaviors will be
likely and when
they will be
effective.

ofpossible
Thesecondareais
pureformsofbehavior UseMultipleTheories
describedinthis
Arecent
articletoabroader
quently
rangeofcontract
Theyargued
altematives.Most
single
research(e.g.,
viewofhumannature,
theauthors
Anderson,1985;
nizational
Eisenhardt,1985,
yieldsamorerealisticview
1988)treatscontracts
Consistent
asadichotomy:
therecommendation
behaviorversus
orywithcomplementary
outcome.However,
orypresents
althoughitisvalid,
contractscanvaryon
acontinuumbetween complexity
spectives
behaviorandoutcome
plexity.
contracts.Also,
Thispointisdemonstrated
currentresearch
pirical
the Singh and
focusesonasingle
Harianto (in press)
reward,neglecting
manysituationsin
whichthereare
multiplerewards,
differingbytime

and Kosnik (1987)


studies support
agency

theory
hypotheses,butthey

frameandcontractba also use the


complementary per
sis.Forexample,upperlevel

spectives

of
are
such
hegemony

and
Bothmultipleandmixedrewards(behaviorand
managerialism.
outcome) present
These perspectives
empirical
difficulties,
but emphasizethepower
they
andpoliticalaspects
alsomirrorreallife.Therichness
ofgoldenparachutes
ityofagency
andgreen
searchers

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions

mail,respectively.Similarly,thestudiesby
Eisenhardt(1988)andConlonandParks
(1988)combineinstitutionalandagency

theories.Theinstitutionalemphasison
traditioncomplementstheefficiency
emphasisofagencytheory,andtheresultisa
betterunderstandingofcompensation.
OtherexamplesincludeAnderson(1985),
whocoupledagencyandtransactioncost,
andEccles(1985),whocombinedagencywith
equitytheory.

LookBeyondEconomics
The
tionalresearchers
nomics
arecarefuldevelopment
logical
ever,muchofthis
menthasalreadybeenaccomplished
theory.Fororganizational
offnow
zational researchers have comparative
advantage(Hirschetal.,1987).Torelytoo
heavilyon
economics
asefficient
styleistoriskdoingsecondrateeconomics
withoutcontributingfirstrate
organizationalresearch.Therefore,
althoughitisappropriatetomonitor
developmentsineconomics,itismoreuseful
totreateconomicsasanadjuncttomore
mainstreamempiricalworkby
organizationalscholars.

Conclusion
Thispaper
onagency
theoryisrevolutionaryand
tion(Jensen,1983)andtheotherarguing
theory
lacks
(Perrow,1986).Amorevalidperspective
themiddle.
realistic,

onproblems
thispaper
rounding
tional
study
suesfacing

References
Anderson,E.(1985)Thesalesperson
ployee:Atransactioncostanalysis.
234254.
Amihud,
motiveforconglomerate
ics,12,605616.
Argawal,
andcorporateinvestmentandfinancingdecisions.
JournalofFinance,42,823837.
Arrow,K.(1971)Essaysinthetheoryofriskbearing.
Chicago:Markham.
Barnard,C.(1938)Thefunctionsoftheexecutive.Cam
bridge,MA:HarvardUniversityPress.
Barney, J. (1988) Agency theory, employee stock
ownership and a firm's cost of equity capital.
Unpublishedworkingpaper,TexasA&MUniversity,
CollegeStation.
Barney,J.,&Ouchi,W.(Eds.)(1986)Organizational
economics.SanFrancisco:JosseyBass.
Basu,A.,Lal,R.,Srinivasan,V.,&Staelin,R.(1985)
Salesforcecompensationplans:Anagencytheoretic
perspective.MarketingScience,4,267291.
Berle,A.,&Means,G.(1932)Themoderncorporation
andprivateproperty.NewYork:Macmillan.
Bolton,M.(1988)Organizationalmiming:Whendolate
adopters of organizational innovations outperform
pioneers? Paper presented at the meeting of the
AcademyofManagement,Anaheim,CA.
Burt,R.(1979)Astructuraltheoryofinterlocking
corporatedirectorates.SocialNetworks,1,415435.
Capitalism in the making. (1984,April 30)Time, p.
62.
Chandler,A.(1962)Strategyandstructure.NewYork:
Doubleday.
Conlon,E.,&Parks,J.(1988)Theeffectsofmonitoring
traditiononcompensation
onprincipal/agentdyads.

72

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions
Eisenhardt, K. (1985)
ceedings(pp.191195).
Anaheim,CA:

Control:

AcademyofMan

Organizational and

agement.

economic approaches.

Cyert, R., & March, J.


(1963) A behavioral
theory of the firm.

Management Science,
31,134149.
Eisenhardt, K. (1988)

Englewood Cliffs, NJ:

Agency

PrenticeHall.

institutional

Dejong,D.,Forsythe,R.,
& Uecker, W. (1985)
Ripoffs, lemons and
reputation formation
in

agency

relationships: A lab

and

explanations

of

compensation

in

retail sales. Academy


of

Management

Journal,31,488511.
Fama,E.(1980)Agency

oratory market study.

problemsandthe

JournalofFinance,50,

theoryofthefirm.

809820.

JournalofPolitical

Demski,J.(1980)A

Economy,88,288307.

simplecaseof

Fama, E., & Jensen, M.

indeterminate

(1983) Separation of

financialreporting.

ownershipandcontrol.

Workingpaper,

Journal of Law and

StanfordUniversity.

Economics, 26, 301

Demski, J.,&Feltham,
G.
(1978)Economic
incentives
inbud

325.
Galbraith, J. (1973)
Designing complex

organizations. Read
getarycontrolsystems.AccountingReview,
ing, MA: Addison
Dornbusch,S.,&Scott,W.R.(1974)Evaluation
Wesley.
ercise
Gardner, W., &
Eccles,
J. Pratt & R.

Martinko, M. (1988)

Zeckhauser (Eds.),

Impression manage

Principals and agents:

ment:

The structure of

observational study

business(pp.151186).

linking

Boston:HarvardBusi

characteristics with

nessSchoolPress.

verbal

An
audience

self

presentations.
Academy of Manage

ment Journal, 31, 42


65.

Holmstrom, B. (1979)
Moral hazard and

Gausch,J.,&Weiss,A.

observability.

Bell

(1981)Selfselectionin

Journal of Economics,

thelabormarket.

10,7491.

AmericanEconomic

Jensen, M. (1983)

Review,71,275284.

Organization theory

Harris,M.,&Raviv,A.

and methodology. Ac

(1978)Someresultson

counting Review, 56,

incentivecontracts

319338.

withapplicationto
educationand
employment,health

Jensen, M. (1984)
Takeovers: Folklore

insurance,andlaw

and science. Harvard

enforcement.

BusinessReview,62(6),

AmericanEconomic

109121.

Review,68,2030.
Harris,M.,&Raviv,A.
(1979)

Jensen, M., & Meckling,


W. (1976) Theory of

Optimal

the firm: Managerial

incentive contracts

behavior,agencycosts,

with

and

imperfect

ownership

information. Journal

structure. Journal of

of Economic Theory,

FinancialEconomics,3,

20,231259.
Hirsch,P.,&Friedman,

305360.
Jensen, M., & Roeback,

R.(1986)Collaboration

R. (1983) The market

orparadigmshift?

for corporate control:

Economicvs.

Empirical evidence.

behavioralthinking

Journal of Financial

aboutpolicy?InJ.
Pearce&R.Robinson
(Eds.),Bestpapers
proceedings(pp.31
35).Chicago:Academy
ofManagement.
Hirsch, P., Michaels, S.,

Economics,11,550.
Kahneman, D., &
Tversky, A. (1979)
Prospect theory: An
analysis of decisions
under

risk.

&Friedman,R.(1987)

Econometrica,47,263

"Dirty hands" versus

291.

"clean models": Is
sociologyindanger of
being seduced by
economics? Theory
andSociety,317336.

Kosnik, R. (1987)
Greenmail:
A
study
in
board
performance
in corporate
governance.
Administrative
Science
Quar
terly, 32, 163185.

Lambert,R.(1983)
Longtermcontracts
andmoralhazard.
BellJournalof
Economics,14,441
452.
Lawrence, P., &Lorsch,
J.(1967)Organization
and
environ
ment.Boston:Divisionof
School.

Judgment calls in
research.BeverlyHills,
CA:Sage.
MeetMikeJensen,the
February
Mitnick,B.(1986)Thetheoryofagency
analysis.
burgh.
Ouchi, W. (1979) A
conceptual framework
for the design of

organizational control
Leatherwood,
mechanisms.
Effectsonpersistenceinaproject.AcademyofManage
mentJournal,30,836848.
Management Science,
MacCrimmon,
management

25,833848.

March,J.(1962)The
businessfirmasa

Pearce,J.,Stevenson,W.,
& Perry, J. (1985)
Managerial

com

politicalcoalition.

pensation based on

JournalofPolitics,24,

organizational

662678.

performance: A time

March,J.,
riskand

series analysis of the

March,J.,&Simon,
Wiley.

Academy

McGrath, J., Martin, J.,

28,261278.

effects of merit pay.

of

Management journal,

& Kukla, R. (1982)

73

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions
Marshfield,MA:Pitt
Perrow, C. (1986)
man.

Complex
organizations. New
York:

Random

House.
Pettigrew,A.(1973)The
politicsof
organizational
decisionmaking.
London:Tavistock.
Pfeffer,J.(1981)Power
inorganizations.

Pfeffer,J.,&Salancik,G.
(1974) Organizational
decision making as a
political process: The
case of a university
budget.Administrative
Science Quarterly, 19,
135151.
Ross,S.(1973)The
economictheoryof
agency:Theprinci

pal'sproblem.

Ungson,G.,&Steers,R.

AmericanEconomic

(1984) Motivation and

Review,63,134139.

politics in executive

Shavell, S. (1979) Risk


sharingandincentives
in the principal and
agentrelationship.Bell
Journal of Economics,
10,5373.

of

Management Review,
9,313323.
Walker,G.,&Weber,D.(1984)Atransaction
tomakeorbuydecisions.Administrative
terly,29,373391.

(1984)Agencytheory,
managerialwelfare,

lationships, takeover
riskand
the
adoption
ofgolden
para

andtakeoverbid
resistance.TheRand
JournalofEconomics,
15,5468.
White,H.(1985)Agency

chutes: An empirical
investigation.
Academy
ofManage

as control. In J. Pratt
& R. Zeckhauser
(Eds.), Principals and

ment Journal.
Sitkin, S. (1987) Secrecy
in organizations: The
limits of legitimate
information control.

Academy

Walking,R.,&Long,M.

Singh, H., &Harianto,


F. (inpress)
Management
board re

Working

compensation.

paper,

University of Texas,

agents: The structure


of business (pp. 187
214). Boston: Harvard
BusinessSchoolPress.
Williamson,0.(1975)
Marketsand

Austin.
Spence,A.M.(1975)The

hierarchies:Analysis

economicsofinternal

andantitrust

organization:An

implications.New

introduction.Bell

York:FreePress.

JournalofEconomics,
6,163172.

Wilson,R.(1968)Onthe
theoryofsyndicates.

Spence,A.M.,&Zeckhauser,R.(1971)Insurance,
Econometrica,36,
tion,andindividualaction.AmericanEconomic
119132.
61,380387.
Thompson, J. (1967)
Organizations

in

action.NewYork:Mc
GrawHill.

Wolfson,M.(1985)Empiricalevidence
andtheirmitigation
Pratt&R.Zeckhauser(Eds.),Principalsandagents:
structureofbusiness
nessSchoolPress.
Zenger,T.(1988)Agency
sorting,agentsolutions
and diseconomies of
scale: An empirical
investigation

of

employment contracts

in high technology

Academy

R&D.Paperpresented

Management,

at the meeting of the

Anaheim,CA.

of

KathleenM.Eisenhardt(Ph.D.,StanfordUniversity)is
Assistant Professor at StanfordUniversity.Correspon
dencecanbesenttoherattheDepartmentofIndustrial
EngineeringandEngineeringManagement,346Terman
Building,StanfordUniversity,Stanford,CA94305.
Theauthor thanksPaulAdler,MicheleBolton,Philip
Bromiley, Jim Hodder, William Ouchi, Gerald Salan
cik, Kaye Schoonhoven, and Robert Sutton for their
commentsandsuggestions.

74

Thiscontentdownloadedfrom114.79.47.36onMon,07Sep201513:38:52
UTCAllusesubjecttoJSTORTermsandConditions