® Academy oí Management Review

1997, Vol. 22, No. 1, 20-47.

TOWARD A STEWARDSHIP THEORY

OF MANAGEMENT
JAMES H. DAVIS
University of Notre Dame
F. DAVID SCHOORMAN
Purdue University
LEX DONALDSON
University of New South Wales

Recent thinking about top management has been influenced by alter-

native models of man.* Economic approaches to governance such as
agency theory tend to assume some form of homo-economicus, which
depict subordinates as individualistic, opportunistic, and self-
serving. Alternatively, sociological and psychological approaches to
governance such as stewardship theory depict subordinates as col-
lectivists, pro-organizational, and trustworthy. Through this research,
.:^we attempt to reconcile the differences between these assumptions by
proposing a model based upon the subordinate's psychological attrib-
utes and the organization's situational characteristics.

Organizatipn_th.ejory and business policy have been strongly influ-

enced by agency theory, which depicts top managers in the large modern
corporation as agents whose interests may diverge from those of their
principals, the shareholders where both parties are utility maximizers
(Jensen & Meckling, 1976). According to agency theory, losses to the prin-
cipal resulting from interest divergence may be curbed by imposing con-
trol structures upon the agent. Although agency theory appears to be the
dominant paradigm underlying most goveirñaHce research and prescrip-
tions, researchers in psychology and sociology have suggested theoreti-
cal limits of agency theory (Hirsch, Michaels, & Friedman, 1987; Perrow,
1986). In particular, assumptions made in agency theory about individu-
alistic utility motivations resulting in principal-agent interest divergence
may not hold for all managers. Therefore, exclusive reliance upon agency
theory is undesirable because the complexities of organizational life are
ignored. Additional theory is needed to explain relationships based upon
other, noneconomic assumptions (Doucouliagos, 1994).
Although agency theory addresses manager-principal interest diver-
gence, additional theory is needed to explain what, if anything, causes

We thank Edward Conlon, Robert Vecchio, Robert House, and Robert Wood for their
helpful comments during the preparation of this article.
' By man we mean the nongender-specific refrence to human beings in general.

20
1997 Davis, Schoorman, and Donaldson 21

interests to be aligned. Stewardship theory has been introduced as a

means of defining relationships basediipon other behavioral premises
(Donaldson & Davis, 1989, 1991). Stewardship theory defines situations
in which managers are not motivated by individual goals, but rather
are stewards whose motives are aligned with the objectives of their
principals. Because stewardship theory is relatively new, its theoretic
contribution has not been adequately established. Previously, research-
ers have contrasted agency and stewardship theories (e.g., Donaldson
& Davis, 1989, 1991, 1994; Fox & Hamilton, 1994), but failed to examine
the psychological and situational underpinnings of stewardship
theory. Clear understanding of the characteristics of the manager and of
the situation are essential to understanding manager-principal interest
convergence. Although the assumptions underlying stewardship theory
have been discussed in general terms (e.g., Donaldson, 1990), as yet,
no author has attempted to define the theory of stewardship in terms
of its underlying assumptions and mechanisms. Finally, previous re-
search seems to be based upon one-best-way thinking, that is, steward-
ship theory is correct and agency theory is incorrect (Donaldson &
Davis, 1991). Research is needed that shows where stewardship theory
fits in the theoretic landscape, relative to agency theory, rather than op-
posed to it.
In this study, we make three contributions to previous stewardship
research. First, we provide a much more detailed description of steward-
ship theory, its language, definitions of terms, and units of analysis. Sec-
ond, we explore the psychological and situational mechanisms that mo-
tivate stewards to behave pro-organizationally. Finally, we do not assume
that agency theory is wrong or inferior to stewardship theory, as previous
researchers have stated. We attempt to reconcile the differences between
stewardship and agency by describing the conditions under which each
is necessary. By articulating stewardship theory in contrast with agency
theory, the boundaries within which each of these two theories applies
can be charted. In these ways, we hope to contribute to the growing body
of stewardship research.
The relationship between stockholders and the manager of a firm has
been described as the "pure agency relationship," because it is associ-
ated with the separation of ownership and control (Jensen & Meckling,
1976). Therefore, in this research, we focus primarily on upper level man-
agers. We begin with a brief description of agency theory, its origins,
underlying assumptions, and theoretical limits. Stewardship theory will
then be described along with its teminology, scope, assumptions, and
limits. The psychological and situational factors that explain manager-
principal interest alignment are discussed. A framework for agency and
stewardship theories is provided, through which the interaction of per-
sonal characteristics of the manager and the characteristics of the situa-
tion are discussed. Through this framework we suggest the theoretical
limits of agency and stewardship and avenues for future research.
22 Academy of Management Review January

AGENCY THEORY

At the heart of agency theory are assumptions of man that can be

traced to 200 years of economic research. The model of man underlying
agency theory is that of a rational actor who seeks to maximize his or her
individual utility (Jensen & Meckling, 1976). Both agents and principals in
agency theory seek to receive as much possible utility with the least
possible expenditure. Thus, given the choice between two alternatives,
the rational agent or principal will choose the option that increases his or
her individual utility.
The advent of the modern corporation created a separation between
ownership and control of wealth (Berle & Means, 1932). Even though own-
ers would prefer to manage their own companies and reap the maximum
utility for themselves, this is impossible because of the capital require-
ments of the modern corporation (Berle & Means, 1932). Corporations grow
beyond the means of a single owner, who is incapable of meeting the
increased economic obligations of the firm. As a result, the modern cor-
poration typically has multiple owners, each intent on maximizing his or
her investment in the enterprise.
Owners become principáis when they contract with executives to
manage their firms for them. As an agent of the principals, an executive
is morally responsible to maximize shareholder utility; however, execu-
tives accept agent status because they perceive the opportunity to maxi-
mize their own utility. Thus, in the modern corporation, agents and prin-
cipals are motivated by opportunities for their own personal gain. Prin-
cipals invest their wealth in companies and design governance systems
in ways that maximize their utility. Agents accept the responsibility of
managing a principal's investments (wealth), because they perceive the
possibility of gaining more utility with this opportunity than by accepting
other opportunities.
If the utility functions of self-serving agents and principals coincide,
there is no agency problem; both agents and principals enjoy increases in
their individual utility. Agency costs are incurred by the principals when
the interests of principals and agents diverge, because given the oppor-
tunity, agents will rationally maximize their own utility at the expense of
their principals. The chance that agents do not share the same interests
and utility choices as their principals is substantial. According to agency
theory, it is difficult for principals to know ex ante which agents will
self-aggrandize, and so it is prudent for the principals to limit potential
losses to their utility (Williamson, 1985). The objective in agency theory
then is to reduce the agency costs incurred by principals by imposing
internal controls to keep the agent's self-serving behavior in check
(Jensen & Meckling, 1976).
Walsh and Seward (1990: 444) argued that "if a firm's managers en-
trench themselves with the sole objective of ensuring their own power,
prestige, and perquisites, the organization is likely to lose sight of its
1997 Davis, Schoorman, and Donaldson 23

compefitive environmental position and will fail." If fhe internal control

mechanisms suggested by agency theorists fail, more expensive, exfernal
control mechanisms (e.g., acquisitions, divestitures, and ownership
amendments) will emerge to control self-serving managers (Walsh &
Seward, 1990). Because of the expense of the external mechanisms to the
principal's utility, internal mechanisms are generally preferred (Walsh &
Seward, 1990).
To profecf shareholder interests, minimize agency costs and ensure
agent-principal interest alignment, agency theorists prescribe various
governance mechanisms. Two mechanisms that have received substan-
tial literary attention are alternative executive compensation schemes
and governance structures (e.g., Demsetz & Lehn, 1985; Jensen & Meckling,
1976). Financial incentive schemes provide rewards and punishments that
are aimed at aligning principal-agent interests. If managers receive com-
pensation that is subject to the successful completion of shareholder ob-
jectives (e.g., long-term rewards tied to firm performance), fhey will be
motivated to behave in a manner consistent with stockholders' interests.
Such incentive schemes are particularly desirable when the agent has a
significant informafional advantage and monitoring is impossible. A sec-
ond mechanism aimed at bringing agents' behavior into alignment with
their principals' interests is governance structure. Boards of directors
keep potentially self-serving managers in check by performing audits and
performance evaluafions. Boards communicafe shareholders' objecfives
and interests to managers and monitor them to keep agency costs in
check. Outside (nonmanagement) board leadership and membership are
desirable to ensure that proper management oversight occurs. Control-
ling governance mechanisms are prescribed, because agency theorists
assume that agent-principal interests may diverge and that given the
opportunity the agent will maximize his or her individual utility at the
expense of fhe principal's utility. Although the divergence of interests
between the agent and principal may differ fo varying degrees, the model
of the agent remains as inherently opportunistic, in that there is an ever-
present possibility of opportunism, unless it is curbed through controls;
moreover, because controls are imperfect, some opportunism will remain.
Agency theorists in no sense specify fotal control of the agent. If
control were total, then the agent would have no discretion and the firm
would be owner-managed. The crux of agency theory is that principals
delegate authority to agents to act on their behalf. If is fhis delegation that
allows agents to opportunistically build their own utility at the expense of
the principals' utility (wealth). Thus, agency theorists specify an interme-
diate condition of confrol, fhat is, first delegation and then controls to
minimize the potential abuse of fhe delegation (Jensen & Meckling, 1976).
The application of agency control does not imply that all managers'
decisions will result in increased wealth for principals; if implies only
thaf the managers will strive to attain outcomes favorable for the prin-
cipals. There are many reasons other than poor motivation for agents'
24 Academy oí Management Review January

failing to deliver high performance for their principals (e.g., low ability,
lack of knowledge, and poor information). Agency theorists are not as
concerned with these failings as they are with those resulting from mo-
tivational problems.
The limits and boundaries of agency theory are determined by its
model of man. Where individualistic, self-serving executive motivation is
assumed, shareholders desirous of minimizing the risks associated with
perceived nonalignment of principal-agent utility functions should imple-
ment agency prescriptions. However, this model has its critics. Jensen and
Meckling (1994) criticized this model of man as being a simplification for
mathematical modeling and an unrealistic description of human behav-
ior. Doucouliagos (1994) argued that labeling all motivation as self-
serving does not explain the complexity of human action. Frank (1994)
suggested that this model of man does not suit the demands of a social
existence. Hirsch, Michaels, and Friedman (1987) said that in exchange for
simplicity and elegance in their models, economists engage in a some-
what broad-brush approach that may reduce empirical verisimilitude and
engender less than robust policies. In short, agency theory assumptions
limit its generalizability.
Agency theory provides a useful way of explaining relationships
where the parties' interests are at odds and can be brought more into
alignment through proper monitoring and a well-planned compensation
system. Additional theory is needed to explain other types of human be-
havior, and this is found in literature beyond the economic perspective. To
that end, stewardship theory will now be described.

STEWARDSHIP THEORY

Stewardship theory has its roots in psychology and sociology and

was designed for researchers to examine situations in which executives
as stewards are motivated to act in the best interests of their principals
(Donaldson & Davis, 1989, 1991). In stewardship theory, the model of man
is based on a steward whose behavior is ordered such that pro-
organizational, collectivistic behaviors have higher utility than individu-
alistic, self:serving behaviors. Given a choice between self-serving be-
havior and pro-organizational behavior, a steward's behavior will not
depart from the interests of his or her organization. A steward will not
substitute or trade self-serving behaviors for cooperative behaviors. Thus,
even where the interests of the steward and the principal are not aligned,
the steward places higher value on cooperation than defection (terms
found in game theory). Because the steward perceives greater utility in
cooperative behavior and behaves accordingly, his or her behavior can be
considered rational.
According to stewardship theory, the behavior of the steward is col-
lective, because the steward seeks to attain the objectives of the organi-
zation (e.g., sales growth or profitability). This behavior in turn will benefit
1997 Davis, Schoorman, and Donaldson 25

principals such as outside owners (through positive effects of profits on

dividends and share prices) and also principals who are managerial su-
perordinates, because their objectives are furthered by the steward. Stew-
ardship theorists assume a strong relationship between the success of the
organization and the principal's satisfaction. A steward protects and
maximizes shareholders' wealth through firm performance, because, by
so doing, the steward's utility functions are maximized.
Given the potential multiplicity of shareholders' objectives, a stew-
ard's behavior can be considered organizationally centered. Stewards in
loosely coupled, heterogeneous organizations with competing stakehold-
ers and competing shareholder objectives are motivated to make deci-
sions that they perceive are in the best interests of the group. Even in the
most politically charged environment, one can assume that most parties
desire a viable, successful enterprise. A steward who successfully im-
proves the performance of the organization generally satisfies most
groups, because most stakeholder groups have interests that are well
served by increasing organizational wealth. Therefore, a pro-organi-
zational steward is motivated to maximize organizational performance,
thereby satisfying the competing interests of shareholders.
This explanation does not imply that stewards do not have necessary
"survival" needs. Clearly, the steward must have an income to survive.
The difference between the agent and the principal is how these needs
are met. The steward realizes the trade-off between personal needs and
organizational objectives and believes that by working toward organiza-
tional, collective ends, personal needs are met. Hence, the steward's op-
portunity set is constrained by the perception that the utility gained from
pro-organizational behavior is higher than the utility that can be gained
through individualistic, self-serving behavior. Stewards believe their in-
terests are aligned with that of the corporation and its owners. Thus, the
steward's interests and utility motivations are directed to organizational
rather than personal objectives.
Stewardship theorists argue that the performance of a steward is
affected by whether the structural situation in which he or she is located
facilitates effective action. If the executive's motivations fit the model of
man underlying stewardship theory, empowering governance structures
and mechanisms are appropriate. Thus, a steward's autonomy should be
deliberately extended to maximize the benefits of a steward, because he
or she can be trusted. In this case, the amount of resources that are nec-
essary to guarantee pro-organizational behavior from an individualistic
agent (i.e., monitoring and incentive or bonding costs) are diminished,
because a steward is motivated to behave in ways that are consistent
with organizational objectives. Indeed, control can be potentially coun-
terproductive, because it undermines the pro-organizational behavior of
the steward, by lowering his or her motivation (Argyris, 1964). The essen-
tial assumption underlying the prescriptions of stewardship theory is that
26 Academy oí Management Review January

the behaviors of the executive are aligned with the interests of the prin-
cipals.
Previously, stewardship theorists have focused on enabling struc-
tures for upper managers (Donaldson & Davis, 1989, 1991, 1994; Fox &
Hamilton, 1994). For example, Donaldson and Davis (1991) argued that, for
CEOs who are stewards, their pro-organizational actions are best facili-
tated when the corporate governance structures give them high authority
and discretion. Structurally, this situation is attained more readily if the
CEO chairs the board of directors. Such a structure would be viewed as
dysfunctional under the agency theory model of man. However, under the
stewardship model of man, stewards maximize their utility as they
achieve organizational rather than self-serving objectives. The CEO-chair
is unambiguously responsible for the fate of the corporation and has the
power to determine strategy without fear of countermand by an outside
chair of the board. Thus, stewardship theorists focus on structures that
facilitate and empower rather than those that monitor and control.
Given the advantage of stewardship to principals, why isn't there
always a steward relationship, rather than an agency relationship? The
answer lies in the risks that the principals are willing to assume. In the
governance contract between owners and executives, owners must decide
how much risk they are willing to assume with their wealth. Risk-averse
owners will most likely perceive that executives are self-serving and will
prefer agency governance prescriptions. Implementing stewardship gov-
ernance mechanisms for an agent would be analogous to turning the hen
house over to the fox. Agency prescriptions can be viewed as the neces-
sary costs of insuring principal utility against the risks of executive op-
portunism. From this perspective, a better question might be Why would
an owner ever take the risks of stewardship governance prescriptions?
Previously, empirical researchers have attempted to validate either
agency theory or stewardship theory as a "one best way" to corporate
governance, assuming that all managers are either stewards or agents.
The results of these studies have resulted in mixed findings; thus, there is
the need for both agency theory and stewardship theory explanations of
management (Donaldson & Davis, 1994). For example, several researchers
found that the agency prescription of independent board leadership (i.e.,
a nonexecutive board chair) is associated with higher firm performance
(e.g.. Berg & Smith, 1978; Daily & Dalton, 1994; Rechner & Dalton, 1991).
Other researchers found that stewardship's executive-chaired boards
have significantly higher corporate performance (e.g., Donaldson & Davis,
1989, 1991; Finkelstein & D'Aveni, 1994). Still others suggest there is no
significant difference in firm performance between executive- and out-
sider-chaired boards (e.g., Chaganti, Mahajan & Sharma, 1985; Molz,
1988). Empirical evidence is similarly mixed with respect to other gover-
nance dimensions (Donaldson & Davis, 1994). The mixed support for
agency and stewardship theories suggests a need to reconcile these
1997 Davis, Schoolman, and Donaldson 27

differences. To this end we now move to a discussion of the situational

and psychological mechanisms underlying the agency and stewardship
models of man.
FACTORS THAT DIFFERENTIATE BETWEEN AGENCY AND
STEWARDSHIP THEORIES
There are a number of dimensions on which agency theory assump-
tions differ from the assumptions of stewardship theory and thereby serve
to differentiate the theories. These dimensions can be characterized
broadly as either psychological factors or situational factors, and they are
discussed in this section.
Psychological Factors
The fundamental difference between agency and stewardship theo-
ries with respect to psychological factors can be traced to the historical
debates regarding the "model of man" described previously in this article.
According to agency theory, man is rooted in economic rationality. In an
interesting response to the work of Simon (1957a,b), Argyris (1973a: 253)
challenged this view of economic man as simplistic regarding human
behavior and argued for a "more complex and humanistic model of man"
in order to incresae the explanatory power and relevance of organization-
al theory. The model of man advocated by Argyris, characterized as "self-
actualizing man," has its roots in the early work of McGregor (1960) and
the later work of Maslow (1970). This model is based on the view that
humans have a need to grow beyond their current state and reach higher
levels of achievement and that the assumptions of the economic view of
man limit people from attaining their full potential. Argyris argued that
when humans are placed in organizations that are designed on this eco-
nomic view, they tend to suppress their level of aspirations, thereby cre-
ating a self-fulfilling prophecy. He further argued that for those individu-
als who are unable to suppress their aspirations, frustration with the
organizational structures may lead to withdrawal and aggressive behav-
iors. The model of man described Argyris is essentially the model of
stewardship theory, and many of the predictions regarding the differ-
ences in the two theories of governance can be traced back to the basic
arguments of the Simon-Argyris debate.
In this article, we focus on the specific differences that are most rel-
evant to the distinctions between agency and stewardship theories. These
differences are reflected in the assumptions about motivation, identifica-
tion, and use of power in the context of the hierarchical relationship ad-
dressed in both theories.
Motivation
The major distinction between agency and stewardship theories is
the focus on extrinsic versus intrinsic motivation. In agency theory, the
focus is on extrinsic rewards: tangible, exchangeable commodities that
28 Academy of Management Review January

have a measurable "market" value. These extrinsic rewards form fhe ba-
sis for fhe reward sysfems that represent the control mechanisms of
agency theory. For example, a principal may create a piecework incentive
system to protect him- or herself from a self-serving agent. Similarly,
medical insurance, 401k savings, and retirement plans may be instituted
as a control mechanism to reduce the likelihood of furnover. Each of fhese
rewards has a quanfifiable value in ferms of dollars fhat is recognized by
both parties. In contrast, in stewardship theory, the focus is on intrinsic
rewards that are not easily quantified. These rewards include opportuni-
ties for growth, achievement, affiliation, and self-actualization. Subordi-
nates in a stewardship relationship are reinforced by these intrinsic,
intangible rewards and are motivated to work harder on behalf of fhe
organizafion. The bases for this distinction can be found in most of the
established theories of motivation, but they are particularly apparent in
the need theories. In a stewardship relationship, the focus would be on fhe
higher order needs of Maslow's hierarchy (1970), on Alderfer's growth need
(1972), and on the achievement and affiliation needs of McClelland (1975)
and McGregor (1966).
A related model of worker motivation, the job characteristics model,
was proposed by Hackman and Oldham (1975, 1976, 1980). These authors
argued that three psychological states (experienced meaningfulness of
work, experienced responsibility for outcomes, and knowledge of the ac-
tual results) mediate the relationship between task characteristics and
internal work motivation. In order to facilifate the attainment of fhese
psychological states, they advocated the redesign of jobs fo increase skill
variefy, task identity, task significance, aufonomy, and feedback. All of
fhese factors are relafed fo increasing fhe opporfunify for growfh and
responsibility for the worker. This model of work mofivafion is consisfent
with the assumptions of sfewardship fheory that increasing the internal
work motivation would lead to higher levels of performance as well as
satisfacfion wifh work. It is interesting to note that in their model Hack-
man and Oldham (1975, 1976, 1980) argued that the growth need strength
of the worker is a moderator of fhe effectiveness of fhis model, suggesfing
that there are some workers for whom fhe assumpfions of the stewardship
model may not fif.
In a more recenf approach to the study of infrinsic mofivafion, Manz
(1986, 1990) developed a theory of self-leadership. According to Manz,
"self-leadership is a comprehensive self-influence perspective that con-
cerns leading oneself fowards performance of naturally motivating tasks
as well as managing to do work that must be done but is not naturally
motivating" (1990: 589). Self-efficacy, self-determination, and feelings of
purpose are characferized as being crifical deferminants of infrinsic mo-
fivation. He argued that self-leadership involves a belief in one's work
that extends beyond the formal reward sysfem and relates to the impor-
tance of shared organizational vision. These views are consistent with the
motivational assumptions of sfewardship theory.
1997 Davis, Schooiman. and Donaldson 29

One other group of motivational theories that present a unique per-

spective on the comparison between agency and stewardship assump-
tions is social comparison theories or equity theories (Adams, 1965; Cosier
& Dalton, 1983). Although the basic premise of an exchange agreement
that is a part of equity theory is more reminiscent of the economic view of
man, the distinction between the perspectives is apparent in the social
comparison that is assumed. In agency theory, there is an economic- or
class-related separation between the principal and the agent. In devel-
oping an equitable work arrangement for the agent, the principal consid-
ers the fair market wage for the agent and arranges the compensation
structure accordingly. The comparison for the agent in determining the
"fairness" of the situation is with respect to other agents in similar con-
texts. In stewardship theory, the principal is a part of the collective and
the basis of comparison would include the principal. Thus, according to
stewardship theory, the principal would expect to be accountable to the
collective for his or her contributions as much as the steward would be.
Although the contributions of the principal and the steward may be quali-
tatively different and not easily quantifiable, the comparison and mutual
accountability are expected.
Proposition 1: People who are motivated by higher order
needs are more likely to become stewards in principal-
steward relationships than are people who are not mo-
tivated by higher order needs.
Proposition 2: People who are motivated by intrinsic fac-
tors are more likely to become stewards in principal-
steward relationships than are people who are moti-
vated by extrinsic factors.
Identification
Identification occurs when managers define themselves in terms of
their membership in a particular organization by accepting the organiza-
tion's mission, vision, and objectives (Kelman, 1958; Mael & Ashforth,
1992), producing a satisfying relationship (O'Reilly, 1989; Sussman & Vec-
chio, 1982). Through identification, an organization becomes an extension
of the steward's psychological structure (Brown, 1969). An identifying
manager interprets comments about the organization as referring also to
himself or herself (i.e., he or she takes the comments personally). Identi-
fication allows managers vicariously to take credit for organizational suc-
cesses and to experience frustration for organizational failures (e.g., Katz
& Kahn, 1978; Turner, 1981). Because managers vicariously take credit for
organizational successes, identification can increase the work-related
satisfaction described previously (e.g., Atkinson, 1957).
A number of authors have found that managers who identify with
organizations attribute organizational success to themselves (e.g., Salan-
cik & Meindl, 1984; Staw, McKechnie, & Puffer, 1983), and this attribution
30 Academy ol Management Review January

contributes to the individual's self-image and self-concept (Kelman, 1961;

Sussman & Vecchio, 1982). This view of organizational identification is
consistent with stewardship theory.
In several studies, researchers argued that managers may external-
ize organizational problems to avoid blame (e.g., D'Aveni & MacMillan,
1990; Staw et al. 1983). When managers externalize attribution for organi-
zational shortcomings, they no longer identify with the organization. In
their effort to avoid incriminating evidence, self-serving managers may
make organizational problems worse, because they avoid accepting re-
sponsibility and making decisions that may rectify the problems (D'Aveni
& MacMillan, 1990). This type of manager falls within the domain of
agency theory.
A manager who idenifies with an organization will thereby work
toward the organization's goals, solve its problems, and overcome barri-
ers that are preventing the successful completion of tasks and assign-
ments (Bass, 1960). When individuals identify with their organizations,
they more readily engage in cooperative, altruistic, and spontaneous un-
rewarded citizenship behaviors (e.g., Mowday, Porter, & Steers, 1982;
O'Reilly & Chatman, 1986; Smith, Organ, & Near, 1983). Therefore, man-
agers who identify with their organization are motivated to help it suc-
ceed and should be empowered to perform their jobs because this will
enable them to use their initiative to promote the success of their organi-
zation and their principals.
A concept that is closely related to identification is organizational
commitment. Porter, Steers, Mowday, and Boulian (1974) defined organi-
zational commitment as the strength of the individual's idenification with
and involvement in a particular organization. They also developed the
organizational commitment questionnaire, which is the most widely used
measure of organizational commitment. In more recent work, Mayer and
Schoorman (1992: 672) characterized organizational commitment as a mul-
tidimensional construct consisting of continuance commitment, which
represents the desire to remain in the organization, and value commit-
ment, which is the "belief in and acceptance of the goals of the organi-
zation." This latter concept of value commitment is more closely related to
the notion of identification, and it is an important component of the psy-
chological profile of a steward. In agency theory, value commitment
would not have economic utility and would not be a relevant part of the
exchange agreement.
Proposition 3; People who have high identification with
the organization aie more likely to become stewards in
principal-steward relationships than are people who
have low identification with the organization.

Proposition 4: People who are high in value commitment

are more likely to become stewards in principal-steward
1997 Davis, Schoorman, and Donaldson 31

relationships than aie people who are low in value com-

mitment.

Use of Power

Power is an important aspect of the relationship between a principal

and a manger. A number of researchers have found that managers re-
ceive satisfaction from, and are motivated by, the use of power (e.g., Mc-
Clelland, 1970, 1975; McClelland & Burnham, 1976). McClelland and Burn-
ham (1976) defined the power motive as a psychological need to influence
others toward the accomplishment of valid and accepted organizational
goals. Managers who have a high need for power tend to "influence or
direct other people; express opinions forcefully; enjoy the role of leader
and may assume it spontaneously" (Steers & Black, 1994: 148).
The types of power used in the context of the relationship help to
differentiate principal-agent relationships from principal-steward rela-
tionships. In the most widely cited typology of power bases, French and
Raven (1959) described power in terms of coercive, legitimate, reward,
expert, and referent power. In a compatible but simpler typology, the five
bases of power are reduced to institutional or organizational power and
personal power (Gibson, Ivancevich, & Donnelly, 1991). Institutional
power is defined as being vested in the principal by virtue of his or her
position in the organization. Thus, termination of organizational member-
ship would terminate the individual's power. The coercive, legitimate,
and many aspects of reward power described by French and Raven (1959)
could be characterized as institutional power. In agency theory, institu-
tional power is the basis of influence in the context of the principal-agent
relationship. In this theory, reward power and legitimate power are used.
Appropriate incentive systems and recognition of authority of the princi-
pal are combined to create the required level of control in the relationship.
Coercive power represents the more severe method of agent control and is
often present in a more subtle form through the threat of termination of
employement. Personal power, an inherent part of the individual in the
context of the interpersonal relationship, is not affected by position. Ex-
pert and referent power are characterized as personal power; referent
power works through identification of one person with another person.
Personal power is developed over time in the context of the relationship
and is not affected by the formal roles in the organization. Although
slower to develop, personal power can be sustained over longer periods of
time. Personal power is the basis of influence in a principal-steward re-
lationship. The choice of the type of power used is a function of the per-
sonal characteristics of the individual and the prevailing organizational
culture. Certain organizational cultures facilitate the use of institutional
power and therefore predispose members to principal-agent relation-
ships. These organizational cultures will be considered next as situ-
ational factors.
32 Academy of Management Review January

Proposition 5: People who aie more likely to use pei-

sonal power as a basis ioi iniluencing others aie more
likely to become stewaids in piincipal-stewaid lelation-
ships than aie people who use insfiiufionaJ powei.

Situational Factors
Management Philosophy
In the early debates between Argyris and the advocates of the eco-
nomic model of man, one of the critical points of contention was whether
organizational theory should be focused on descriptive or normative mod-
els of the organization. Simon (1957a,b; 1973) and others (e.g., Cyert &
March, 1963) argued that the economic model, and therefore implicitly the
agency theory assumptions, were the predominant basis of relationships
in organizations. They cited numerous examples of behavior by both prin-
cipals and agents to support this claim. In contrast, Argyris (1973a,b) ar-
gued that the management philosophy of most organizations was based
on economic assumptions and that this became a self-fulfilling prophecy
regarding the nature of relationships that would develop. He advocated
the development of normative models of organization based on self-
actualizing assumptions in order to create an organizational culture that
supported the development of stewardship types of relationships. The
position advocated by Argyris (1973) was similar to the arguments ad-
vanced earlier by McGregor (1960) in his discussion of Theory Y manage-
ment and by Likert (1961) in his comparison of System 4 management with
more control-oriented systems. The common message in each of these
theories is that the assumption about the model of man drives the devel-
opment of management philosophies and management systems, which
then serve to produce behavior in the organization that is consistent with
the assumptions. Each of these theorists advocated the development of
normative models of organization and a break from the traditional man-
agement philosophies in order to facilitate the self-actualizing behaviors
that are consistent with stewardship theory.
More recently, Walton (1980, 1985) advocted what he called a high-
commiímení management philosophy. This approach to management
was characterized as being highly participative and consisting of open
communication, empowerment of workers, and the establishment of
trust. Lawler (1986, 1992) elaborated on this view by contrasting the man-
agement philosophies he described as confroi orienfed versus invoJve-
menf orienfed. According to Lawler, the control-oriented approach is
based on a management philosphy that the thinking and controlling
part of the work must be separated from the doing part of the work.
In contrast, involvement-oriented approaches emphasize self-control
and self-management and do not create a separation among thinking,
controlling, and doing the work. The key assumption in involvement-
oriented approaches is that when employees are given challenges and
1997 Davis, Schooiman, and Donaldson 33

responsibility they will develop self-control of fheir behavior. Lawler

(1986, 1992) characterized the control-oriented approaches as a mature
philosophy of managemenf that flourished in fhe 1960s and 1970s largely
because fhe compefifive advantages of organizations in the United States
were not based on their management philosophies. He characterized the
involvement-oriented approach as a newer approach that is still not as
widely adopted as the control-oriented approach. Although Lawler advo-
cated the adoption of an involvemenf-oriented management philosophy
as the merging dominant approach, he made this argument through a
contingency model. When short-term cost control and productivity are
important issues, the control-oriented approach produces better results.
However, he argued that this approach cannot be sustained in the long
term because of inherently faulty assumptions about the motivation of
workers. Thus, when labor costs are low and unemployement is high, the
control-oriented approach may work well, because turnover due to em-
ployee dissatisfacfion will be minimal and replacemenf costs are low. In
contrast, in an uncertain environment, with high labor costs, a focus on
long-ferm effecfiveness and qualify through self-inspection, or the in-
volvement-oriented approach, has significanf advanfages. The argu-
ments in favor of fhe involvement-oriented approach as the dominant
manaagement philosophy of fhe fufure are based on observafions that the
environment is changing in ways that make control-oriented approaches
less viable.
One important difference between the two management philosophies
is in their orientation to risk. We have already noted that in unstable,
uncertain environments the involvement-oriented approach is more effee-
five, but in stable environments, the control-oriented approach is best.
When control-oriented management encounters an uncertain or risky situ-
ation, it manages the risk through the implementation of greafer controls.
For example, if fhe product design becomes more complicated, the orga-
nization may introduce a quality control unit to inspect the finished parts
for defecfs. As workers feel less mofivated because of boring jobs, more
supervisors would adopt the control-oriented solution. In contrast, in the
involvement-oriented approach, the means of dealing with increased un-
certainty and risk is through more training, empowerment, and ultimately
trust in workers. In the quality example, workers would be given addi-
tional training on the complex product and given the responsibility for
self-inspection of qualify. If fhe jobs are boring, they would be redesigned
to be more challenging and therefore more mofivating.
The issue of frusf is a critical aspect of the high-commitment or in-
volvement-oriented management philosophy. Work by Mayer, Davis, and
Schoorman (1995) defined trust as a willingness to be vulnerable in the
context of a relationship. Control-oriented systems are designed to avoid
vulnerability and therefore to avoid the need to trust. Another important
aspect of frusf is fhat it occurs in the context of a relafionship, and it is
most likely to occur when the relationship is based on personal power
34 Academy oí Management Review January

(respect and expertise). In a control-oriented approach, the relationships

are generally transactional in nature or are based on institutional power.
The key point of this discussion is that the management philosophy of
an organization creates a context in which the choice of agency or stew-
ardship relationships is made by principals and managers. A control-
oriented management philosophy is more likely to produce choices of
agency theory relationships, whereas an involvement-oriented manage-
ment philosophy is more likely to produce stewardship theory relation-
ships. This view is completely consistent with the observations of Argryis
(1973) that the design of organizations based on economic assumptions
creates a self-fulfilling prophecy of producing behavior that is consistent
with the assumptions. If we follow this reasoning, evolution of the involve-
ment-oriented management philosophy into a more dominant model will
lead to the emergence of behavior that is more consistent with steward-
ship theory.
Proposition 6; People who are in an involvement-
oriented situation are more likely to become stewards in
principal-steward relationships than are people who
are in a control-oriented situation.
Culture

Individualism-collectivism. There are also aspects of culture that may

influence the choice between agency and stewardship relationships. In
his pioneering work on cultural differences, Hofstede (1980, 1991) de-
scribed the dimension of individualism-collectivism. Individualism is
characterized as the emphasis of personal goals over group goals. Col-
lectivists subordinate their personal goals to the goals of the collective
(Triandis, 1995; Triandis, Dunnette, & Hough, 1993). Hofstede (1980) found
that nations and regions of the world can be described according to the
orientation on this individualism-collectivism dimension. For example,
individualism is a cultural pattern found in the United States, Canada,
and western Europe. Collectivism is common in Asia, South America, and
southern Europe. Although much of this research was focused on the cul-
tural pattern of a nation, there is distinctive variation within nations
(Triandis, 1995, 1990). The generally accepted view is that the national
culture predisposes members of that culture to either a collectivist or an
individualistic orientation. However, the extent of this influence varies
among individuals, and the effects of other experiences shape the ulti-
mate orientation of each person.
Several specific differences between individualists and collectivists
are relevant to the choice between agency and stewardship theory rela-
tionships. In collectivist cultures, the self is defined as a part of the group.
One's group memberships (e.g., family, university, and organization) are
an important statement of identity and achievement. In collectivist cul-
tures individuals are usually addressed by family names, whereas in
1997 Davis, Schooiman, and Donaldson 35

individualistic cultures "first" or given names are preferred. Success is

defined in terms of the success of the group. Collectivists have a very
positive attitude toward harmony in groups, avoiding conflict and con-
frontation. Individualists see confrontation as an opportunity to "work
things out" and to communicate more directly. Collectivists prefer long-
term relationships and will frequently take a longer time and expend
greater effort to "get to know" someone prior to a business transaction.
The development of the relationship is an important first step in business
dealings, which often depend on a "handshake" or trust. Individualists
are more short-term oriented, conduct business independently of personal
relationships, use a cost-benefit analysis (economic model) to evaluate
the business exchange, and will reduce the risks of doing business by
signing a contract.
It should be apparent from the preceding discussion that collectivist
cultrues are more conducive to the emergence of stewardship relation-
ships and that collectivists are more likely to initiate a principal-steward
relationship. Individualistic cultures would appear to facilitate agency
relationships.
Proposition 7: People in a collectivist culture are more
likely to develop principal-steward relationships than
are people who are in an individualistic culture.
Power distance. A second dimension developed by Hofstede (1980,
1991) to characterize the cross-cultural differences that is particualarly rel-
evant to the agency-stewardship distinction is the concept of power dis-
tance. Power distance is generally defined as "the extent to which less
powerful members of institutions and organizations within a country ex-
pect and accept that power is distributed unequally" (Hofstede, 1991: 28).
According to Hofstede (1980, 1991), in certain cultures, relatively large
differences in power among members are accepted and tolerated more
than they are in other cultures. In a culture with high power distance,
there is an acceptance that less powerful members will be dependent on
more powerful members and privileges and status symbols are both ex-
pected and popular. Class and caste systems are an accepted part of this
culture. In low power distance cultures, inequalities are minimized, inde-
pendence of the less powerful is valued and encouraged, and status and
class symbols are frowned upon (Hodgetts & Luthans, 1993). The concept
of power distance has its roots in the family structure and is pervasive in
the institutions that socialize members of the culture (e.g., school, church,
and social organizations). In high power distance cultures, children are
expected to be obedient to their parents; respect for parents and elders is
considered a basic virtue (Hofstede, 1991). In such cases, children are
"looked after" and allowed to be dependent for a longer period of time,
and, in tum, children treat parents and grandparents with formal defer-
ence even through adulthood. In contrast, in low power distance cultures,
children are treated as equals, are encouraged to be independent at a n .
36 Academy oí Management Review January

early age, and relationships are not related to status or role. Formal re-
spect and deference are seldom shown. This pattern of differences is
also observed in organizational life. In high power distance cultures, or-
ganizations are centralized, and they include large differences in author-
ity, salary, and privileges between those at the top and those at the bot-
tom. In low power distance cultures, organizations are decentralized,
there is more consultation in decision making, and the differences in
salary and perquisites are minimized. Similar to the discussion of the
individualism-collectivism dimension, Hofstede (1980, 1991) and others
(e.g., Triandis, 1995) have argued that although the national culture cre-
ates a predisposition to either high or low power distance, there can be
considerable variance in power distance across organizations and indi-
viduals in the same country.
High power distance cultures are conducive to the development of
agency relationships, because they support and legitimize the inherent
inequality between prinicpal and agent. This idea is especially true in the
context of work, because the development of hierarchies, of layers of su-
pervision (as control mechanisms), and of inequalities in rewards and
status may lead the agents to "ideologically reject the boss's authority
completely, while in practice they will comply" (Hofstede, 1991: 35). This
characterization by Hofstede is similar to the predictions regarding the
self-serving agent described in agency theory. Low power distance cul-
tures are more conducive to the development of stewardship relation-
ships, because their members place greater value on the essential equal-
ity of the principal and the manager. This orientation encourages the
development of relationships between principals and managers that are
an essential part of stewardship theory.
Proposition 8; Peopie in a iowpower distance culture are
more likely to develop principal-steward relationships
than are people who are in a high power distance cul-
ture
Although the individualism-collectivism and power distance dimen-
sions are not perfectly correlated, there appears to be a pattern of rela-
tionships that make the predictions regarding the cultural antecedents of
stewardship theory somewhat complicated. For example, the United
States is generally regarded as an individualistic culture with low power
distance. Individualism would suggest a predisposition to agency theory,
whereas low power distance would predict more stewardship theory out-
comes. Simiarly, Japan is a high power distance, collectivist culture, lead-
ing to similarly conflicting predictions. The apparent contradictions in
predictions suggested by these variables may provide a valuable expla-
nation for the process by which agency and stewardship relationships
develop. For example, we might expect that members of a collectivist
culture would move very quickly to establish an organizational structure
that is conducive to the development of stewardship relationships (e.g..
1997 Davis, Schoolman, and Donaldson 37

flat decentralized, and team based), but they would have difficulty de-
veloping the truly participative, challenging style of trusting interper-
sonal relationships because of high power distance. This is a common
characterization of Japanese attempts to develop high-involvement man-
agement philosophies. In contrast, in the United States, we might expect
to encounter great resistance to the restructuring of organizations into
flatter, more decentralized, team-based units, but once this restructuring
is accomplished, we would expect the members to function well as a
high-involvement team.
The conflict that is suggested by the cultural dimensions is not lim-
ited to culture. It is certainly conceivable that some of the psychological
mechanisms may suggest one theory, whereas other mechanisms may
suggest the alternate. We might also expect the possibility of a mismatch
between the management philosophy and the psychological characteris-
tics of the managers. Although the specific interactions of these anteced-
ents in the prediction of stewardship versus agency relationships is very
intriguing, we believe it is beyond the scope of this article. The further
development of this model in terms of the more specific predictions would
be a logical next step regarding the empirical testing of the main effects
specificed in this model.
The disucssion so far has focused on the core issues underlying
agency and stewardship theories. A summary of the primary differences
between the two theories is shown in Table 1. The primary difference lies

TABLE 1
Comparison of Agency Theory and Stewardship Theory
Agency Theory Stewardship Theory
Model of Man Economic man Self-actualizing man
Behavior Self-serving Collective serving

Psychological Mechanisms
Motivation Lower order/economic Higher order needs (growth,
needs (physiological, achievement,
security, economic) self-actualization)
Extrinsic Intrinsic
Social Comparison Other managers Principal
Identification Low value commitment High value commitment
Power Institutional (legitimate, Personal (expert, referent)
coercive, reward)

Situational Mechanisms
Management Philosophy Conrol oriented Involvement oriented
Risk orientation Control mechanisms Trust
Time frame Short term Long Term
Objective Cost control Performance Enhancement
Cultural Differences Individualism Collectivism
High power distance Low power distance

?)
38 Academy of Management Review January

in the assumptions about human nature. According to agency theory,

people are individualistic, utility maximizers. According to stewardship
theory, people are collective self-acfualizers who achieve utility through
organizational achievement.

THE CHOICE BETWEEN AGENCY AND STEWARDSHIP RELATIONSHIPS

In the preceding sections, we presented a model suggesting that

there are psychological and situational factors that predispose individu-
als to agency and stewardship approaches to relationships. As we dis-
cussed, many authors argured that humans prefer growth, responsibility,
and self-actualization and advocafed an involvemenf-oriented manage-
menf philosophy and frusf as a mechanism for dealing wifh risk. Alfhough
many of these researchers contended that these motivations are univer-
sally shared by all people, we frame these issues as a model in which the
psychological and situational characteristics of the principal and man-
ager are antecedents of their choice between agency and stewardship
relationships. (The position that there is one best choice for psychological
[e.g., self-actualizing assumptions] or situational [e.g., a universally dy-
namic environment] reasons would represent a special case of fhe model.)
The choice beween acfivity and stewardship relationships is similar
to the decision posed by a prisoner's dilemma. First, it is a decision made
by both parties to the relationship. The psychological characteristics of
each party predisposes each individual to make a particular choice. Sec-
ond, the situational characteristics have an influence on the choice. The
management philosophy may have a significanf impact on the choice by
both parties. The cultural background (collectivism and power distance)
of each party will also affecf fhe choice. Finally, fhe expectation that each
party has of fhe other will influence the choice between agency and stew-
ardship. A longer history of fhese parties dealing with each other will
provide more data to guide these expectations.
The nature of fhe dilemma is illustrated in Figure 1. When both the
principal and the manager choose an agency relationship, the result is a
true principal-agency relationship that is likely to achieve the expecta-
tions of each. The agency relationship is designed to minimize potential
losses to each party. The manager's psychological profile fits that of fhe
agent, and, thus, he or she will use any discretion to the disadvantage of
the organization and principals. Such a manager requires a controlling
situation to hold his or her opportunistic tendencies in check. The pres-
ence of confrols in fhis case constitutes a fif and ensures fhat the agency
costs are minimized. Thus, both parties have similar expectations of fhe
relationship, and costs are controlled.
When both the principal and the manager choose a stewardship re-
lationship, the results is a true principal-steward relationship that is de-
signed to maximize the potential performance of the group. In this situa-
tion, the manager has the psychological profile of a steward and thus
1997 Davis, Schooiman, and Donaldson 39

nGURE'l
Principal-Manager Choice Model

Principal's Choice

Agent Steward

Minimize Potential Agent Acts

Costs Opportunistically

Mutual Agency Principal Is Angry

Agent Relationship
Principal Is Betrayed

1 2
Manager's
Choice
3 4

Principal Acts Maximize Potential

Steward Opportunistically Performance

Manager Is Frustrated Mutual Stewardship

Relationship
Manager Is Betrayed

gains utility from fulfilling the purposes and objectives of the organiza-
tion. Likewise, the principal chooses to create a stewardship situation that
is involvement oriented and empowering. The mutual gains resulting
from this state of fit are high.
The dilemma occurs because there is the possibility of a different
choice by each party. If the principal chooses an agency relationship and
the manager chooses a steward relationship, the result is likely to be a
very frustrated manager who feels betrayed by the principal. When stew-
ards are controlled as if they were agents, they cannot enjoy the types of
internal rewards they desire (i.e., growth, achievement, or self-
actualization), and as a reult, they may engage an antiorganizational
behaviors (Argryis, 1964). The application of control may create disenfran-
chised employees, because principals, rather than those persons actually
doing the work, assume the responsibility of deciding and orchestrating
firm procedures. Managers in controlling, less trusting climates may not
have the opportunity to behave as stewards and therefore may experience
decreased feelings of self-worth, self-responsibility, and self-control
(Argryis, 1964) and have less desire to behave as stewards. In this
situation, the workplace becomes depersonalized, and the manager may
begin to view him- or herself as an interchangeable unit. In such envi-
ronments, employees may resort to antagonistic adaptive activities such
40 Academy ot Management Review January

as absenteeism and turnover (Fleishman & Harris, 1962; James, Demaree,

Mulaik, & Ladd, 1992); theft and vandalism; poor workmanship; slow-
downs; stealing; causing waste (Argryis, 1964; James et al., 1992); and
demanding better financial compensation, benefits, and working condi-
tions (Herzberg, Mausner, & Snyderman, 1959).
If the principal chooses a steward relationship and the manager
chooses an agency relationship, the manager acts opportunistically and
takes advantage of the principal. A manager whose psychological profile
fits that of an agent will behave as a "fox in the henhouse" and will seek
to satisfy his or her personal utility at the expense of the organization and
the principal. The situation created by the principal empowers the agent
to work only to serve him- or herself. Thus, the psychological profile of the
manager is out of harmony with the situation created by the principal. The
principal is likely to feel betrayed and angry and may increase controls,
withdraw from the situation, or attempt to remove the manager.
In Table 1, we presented a comparison of the characteristics of
agency and stewardship theories. Because the apaprent dominance of
stewardship theory one would ask why it would not be adopted by ev-
eryone. We believe the answer is in the level of risk that is acceptable to
each individual and his or her willingness to trust the other party. Al-
though the highest joint utility is in the principal-steward relationship, in
which both parties choose the steward relationship (Cell 4), the least risk
of betrayal (losses) is in the principal-agent relationship (Cell 1), in which
both parties choose the agency relationship. It is easy to see from this
illustration that when each party has an individualistic orientation, the
best choice (regardless of the choice of the other person) is an agency
relationship. Thus, when two individualistic parties are involved, the in-
evitable choice is an agency relationship. Only in a collectivist orienta-
tion, when both parties subordinate their personal goals to that of the
collective, will they evaluate the joint utility and mutually choose a stew-
ardship relationship.
Proposition 9: If a mutual stewardship relationship ex-
ists, potential performance of the firm is maximized.
Proposition 10: If a mutual agency relationship exists,
potential costs of the firm are minimized.
Proposition 11: If a mixed-motive choice exists, the party
choosing stewardship is betrayed, and the party choos-
ing activity is opportunistic.
FUTURE RESEARCH
We have attempted to sketch a broad outline of the psychological
and situational processes that are presently somewhat neglected in
contemporary management theory and that provide the underpinnings
of stewardship theory. More fine-grained analyses are needed, which
would include more detailed theory construction, the examination of new
1997 Davis, Schoorman, and Donaldson 41

variables, and empirical testing. In the future, researchers should inquire

into the stewardship mechanisms identified in this article and examine
their relative importance, their interactions, and the situational contin-
genices that affect them.
For our theory, we adopted the simplifying assumption of a choice of
agency versus stewardship relationships at a single point in time (i.e., on
trial one in a relationship). This assumption was necessary as a first step
in establishing the contrast between agency and stewardship theories
and for the development of the framework for the choice model. Although
this simplification could be viewed as a limitation of our theory, the in-
corporation of the dynamic aspects of the theory were determined to be
beyond the scope of this article. Clearly, the role of a long-term relation-
ship is central to the choice of stewardship roles. One of the important
implications of the theory of stewardship presented here is that if a
mixed-motive choice is made and one party is betrayed, the inevitable
progression of the relationship is toward an agency model. Researchers
should explore the choice of agency versus stewardship relationships
over time, incorporating variables that capture the dynamic nature of
principal-manager relationships.
In developing a model of the choice between agency and stewardship
theories, we have specified the psychological and situational antecedents
of the choice in terms of direct main effects. However, we have not ruled
out the possibility that even among the variables identified in this article
there are likely to be more complex interactions and dynamic effects over
time that determine the eventual choice. For example, in an organization
that has a high-involvement culture, managers may change over time and
learn to value the growth opportunities presented by the job; thorugh
increased value commitment and identification, managers may develop
and use more personal power. Thus, the argue that there may be an in-
teraction effect between the organization's philosophy and the psycho-
logical variables, and in a long-term model, there may be a direct effect
of the situation (philosophy) on the psychological factors. This argument
is consistent with Argryis's (1964) views regarding the self-fulfilling
prophecy created by organizational philosophy and structure. Research-
ers should explore these potential interactions.
An interesting implication of the theory is related to the cultural vari-
ables and the process of implementation of structural changes in organi-
zations. As we noted previously, there appear to be several national
cultures in which the values for individualism, collectivism, and power
distance would lead to opposite predictions regarding the propensity for
stewardshi prelationships. A more fine-grained examination of these
variables may suggest that the impact of the cultural variables may be
felt at different stages in the process of implementing structural change.
Consider the example of an organization that is changing its traditional
hierarchical structure to one that is flatter, more decentalized, more partici-
pative, and therefore more conducive to stewardship types of relationships.
42 Academy of Management Review January

(In terms of our model, this would be represented as a change from a

control-oriented philosophy to an involvement-oriented philosophy). In
the United States (high individualism, low power distance), we might
expect this process to meet greater resistance as a firm moves toward
collective action, but once such a structure is in place, the participants
move rapidly to develop a highly participative environment where there
is greater equality. In contrast, in Japan (low individualism, high power
distance) we might expect such changes to proceed more smoothly as
members accept the team concept, but progress beyond that point might
meet great resistance in the absence of "leaders" who provide direction
for the teams. Although these scenarios are speculative, they illustrate
the richness of these issues as avenues for future research.
Another potential area for future research would include the relation-
ships among theories about trust in organizations, risk-taking behavior,
and stewardship. There has been much interest in the development of
models of organizational trust (e.g.. Chiles & McMackin, 1996; Hosmer,
1995; Mayer et al., 1995). It should be clear in this article that stewardship
theory and the choices of stewardship relationships in organizations rely
heavily on the trust between the principal and managers as well as the
perceived risks. As we noted previously, Mayer and colleagues (1995: 712)
defined trust as "the willingness to be vulnerable." This definition of trust
is the antithesis of the basic premise of agency theory, which could be
restated as "the unwillingness to be vulnerable." We believe there is
much to be learned from exploring the relationships among trust and
stewardship in organizations.
Finally, the stewardship theory presented here could be integrated
into contemporary thinking regarding leadership in organizations. Are
charismatic leaders more likely to develop principal-steward relation-
ships? Are transactional leaders following the agency model? Also, is
leadership a dyadic process (Liden, Wayne, & Dean, 1993; Scandura &
Schriesheim, 1994)? When considered in the context of stewardship
theory, this issue has very interesting implications. According to our
theory, the choice of stewardship relationships is made one relationship
at a time, and the success of the relationship is a function of the mutual
choice by two parties in the relationship. This idea implies that any one
principal could have both agency and stewardship relationships with
mutliple manages at the same time and that managers could have both
agency and stewardship relationships with different principals. Each of
these issues deserves further investigation.

CONCLUSIONS

We use agency theory to help researchers to understand the conflicts

of interest that can arise between principals and agents, the result-
ing potential problems of opportunism, and the structures that evolve to
contain it, such as supervision and incentives. However, organizational
1997 Davis, Schoorman, and Donaldson 43

relationships may be more complex than those analyzed through agency

theory. The propositions of agency theory may not apply in all situations.
An alternative model of managerial mofivafion and behavior is steward-
ship theory, which is derived from psychological and sociological fradi-
fions. Our research adds fo the understanding of sfewardship theory by
describig its terminology and theoretical contribution.
We extend previous stewardship theory research by defining several
of fhe psychological and sociological characteristics that are antecedents
to principal-steward relationships. Managers whose needs are based on
growth, achievement, and self-actualization and who are intrinsically mo-
tivated may gain greater utility by accomplishing organizational rather
than personal agendas. Likewise, managers who identify with their or-
ganizations and are highly committed to organizational values are also
more likely to serve organizational ends. Finally, situations in which the
managerial philosophy is based on involvement and trust and the culture
is based on collectivism and low power distance generally result in prin-
cipal-steward relationships.
We also add to previous stewardship research by examining a model
based on manager-principal choice rather than a determinism. Accord-
ing to our model, managers choose to behave as stewards or agents. Their
choice is contingent on their psychological motivations and their percep-
tions of fhe situation. Principals also choose to create an aency or stew-
ardship relationship, depending upon their perceptions of fhe sifuation
and fhe manager. If eifher fhe manager or the principal perceives that the
other party will behave in an activity manner (defect), it is in his or her
best interest to behave in an agency fashion, and fhe organizafion
receives a subopfimal refurn on its investment. If both parties choose to
develop a stewardship relationship (cooperate), the organization realizes
the maximum reward. Unlike previous researchers, who assumed that
managers are predisposed to act like stewards or agents, we base our
research on choice rather than on determinism.
Finally, we suggest future avenues for sfewardship theory research.
We describe a need for more fine-grained analysis of the proposed psy-
chological and situational factors. Through such research, management
scholars can come to a clearer understanding of these and other variables
that may influence principal-manager relationships. We also argue that
more dynamic modeling is necessary to understand how time and prior
decisions affect fufure relafionships. We call for research on the interac-
tions among the psychological mechanisms and situational factors and
the relationship between trust and risk that each party is willing to as-
sume. In short, a variety of theoretical and empirical projects are needed
to help researchers to fully undersfand sfewardship fheory.
REFERENCES
Adams, J. S. 1985. Injustices in social exchange. In L. Berkowitz, (Ed.), Advances in experi-
mental social psychology: 287-299. New York: Academic Press.
44 Academy oí Management Review January

Alderfer, C. P. 1972. Existence, relatedness. and growth: Human needs in organizational set-
tings. New York: Free Press.
Argryis, C. 1964. integrating the individual and the organization. New York: Wiley.
Argryis, C. 1973a. Organization man: Rational and self-actualizing. Public Administration
Review. 33 (July/August): 354-357.
Argryis, C. 1973b. Some limits of rational man organizational theory. Public Administration
fleview, 33 (May/June): 253-267.
Atkinson, J. W. 1957. Motivational determinants of risk-taking behavior. Psychological Re-
view. 64: 359-372.
Bass, B. M. 1960. Leadership, psychology, and organizational behavior. New York: Harper.
Berg, S. V., & Smith, S. K. 1978. CEO and board chairman: A quantitative study of dual vs.
unitary board leadership. Directors & Boards, 3(l):34-39.
Berle, A., & Means, G. 1932. The modern corporation and private property. New York: Mac-
millan.
Brown, M. E. 1969. Identification and some conditions of organizational involvement. Admin-
istrative Science Quarterly, 14: 346-355.
Chaganti, R. S., Mahajan, V., & Sharma, S. 1985. Corporate board size, composition and
corporate failures in retailing industry. Journal oí Management Studies. 22: 400-417.
Chiles T. H., & McMackin, J. F. 1996. Integrating variable risk preferences, trust, and trans-
action cost eocnomics. Academy oí Management Review. 21: 73-99.
Cosier, R. A., & Dalton, D. R. 1983. Equity theory and time: A reformulation. Academy oí
Management Review. 8: 311-319.
Cyert, R. M., & March, J. G. 1963. A behavioral theory ol the ¡irm. Englewood Cliffs, NJ: Pren-
tice Hall.
Daily, C. M., & Dalton, D. R. 1994. Bankruptcy and corporate governance: The impact of board
composition and structure. Academy oí Management Journal. 37: 1603-1617.
D'Aveni, R. A., & MacMillan, I. C. 1990. Crisis and the content of managerial communications:
A study of the focus of attention of top managers in surviving and failing firms. Admin-
istrative Science Quarterly, 35: 634-657.
Demsetz, H., & Lehn, K. 1985. The structure of corporate ownership: Theory and consequencs.
Journal of Political Economics. 93: 11-55.
Donaldson, L. 1990. The ethereal hand: Organizational economics and management theory.
Academy oí Management Review. 15:369-381.
Donaldson, L., & Davis, J. H. 1989. CEO governance and shareholder returns; Agency theory
or stewardship theory. Paper presented at the annual meeting of the Academy of Man-
agement, Washington, DC.
Donaldson, L., & Davis, J. H. 1991. Stewardship theory or agency theory: CEO governance and
shareholder returns. Australian Journal oí Management. 16: 49-64.
Donaldson, L., & Davis, J. H. 1994. Boards and company performance—Research challenges
the conventional wisdom. Corporate Governance: An International Review, 2: 151-160.
Doucouliagos, C. 1994. A note on the volution of homo economicus. Journal o/ Economics
issues, 3: 877-^83.
Finkelstein, S., & D'Aveni, R. A. 1994. CEO duality as a double-edged sword: How boards of
directors balance entrenchment avoidance and unity of comand. Academy ol Manage-
ment Journal. 37: 1079-1108.
1997 Davis, Schooiman, and Donaldson 45

Fleishman, E. A., & Harris, E. F. 1962. Patterns of leadership behavior related to employee
grievances and turnover. Personnel Psychology, 15: 43-56.
Fox, M. A., & Hamilton, R. T. 1994. Ownership and diversification: Agency theory or steward-
ship theory. Journal ol Management Studies, 31:69-81.
Frank, R. H. 1994. Microeconomics and behavior. New York: McGraw-Hill.
French, J. R. P., & Raven, B. 1959. The bases of social power. In D. Cartwright (Ed.), Síudies in
social power: 150-167. Ann Arbor: University of Michigan Institute for Social Research.
Gibson, J. L., Ivancevich, J. M., & Donnelly, J. H. 1991. Organizations. Homewood, IL: Irwin.
Hackman J. R., & Oldham, G. R. 1975. Development of the job diagnostic survery. Journal ol
Applied Psychology, 60: 159-170.
Hackman, J. R., & Oldham, G. R. 1976. Motivation through the design of work: Test of a theory.
Organizational Behavior and Human Performance, 15: 250-279.
Hackman, J. R., & Oldham, G. R. 1980. Work redesign. Reading, MA: Addison-Wesley.
Herzberg, F., Mausner, B., & Snyderman, B. B. 1959. The motivation to work. New York: Wiley.
Hirsch, P., Michaels, S., & Friedman, R. 1987. "Dirty hands" versus "clean models." Theory and
Society, 16: 317-336.
Hodgetts, R. M., & Luthaus, F. 1993. U.S. multinationals' compensation strategies for local
management: Cross-cultural implications. Compensation and Benelits Review, 25:42-48.
Hofstede, G. 1980. Cuifure's consequences: /nternafiona/ dillerences in worJ-reJaied values.
Beverly Hills, CA: Sage.
Hofstede, G. 1991. Cultures and organizations: Soltware ol the mind. London: McGraw-Hill.
Hofstede, G. 1993. Cultural constraints in management theories. Academy ol Management
Executive, 7(1): 81-90.
Hosmer, L. T. 1995. Trust: The connecting link between organizational theory and philosoph-
ical ethics. Academy ol Management Review, 20: 379-403.
James, L. R., Demaree, R. G., Mulaik, S. A., & Ladd, R. T. 1992. Validity generalization in the
context of situational models. Journal ol Applied Psychology, 77: 3-14.
Jensen, M. C, & Meckling, W. H. 1994. The nature of man. Journal ol Applied Corporate
Finance, 7(2): 4-19.
Jensen, M. C, & Meckling, W. H. 1976. Theory of the firm: Managerial behavior, agency costs
and ownership structure. Journal ol Financial Economics, 3: 305-360.
Katz, D., & Kahn, R. L. 1978. The social psychology ot organizations (2nd ed.). New York: Wiley.
Kelman, H. C. 1958. Compliance, identification, and intemalization: Three processes of atti-
tude change. Journal ol Conflict Resolution, 2: 51-60.
Kelman, H. C. 1961. Processes of opinion change. Public Opinion Quarterly, 25: 57-78.
Lawler, E. E. 1986. High involvement managmeent. San Francisco: Jossey-Bass.
Lawler, E. E. 1992. The ultimate advantage. San Francisco: Jossey-Bass.
Liden, R. C, Wayne, S. J., & Dean, S. (1993). A longitudinal study on the early development of
leader-member exchange. Journal ol Appliced Pschology, 78: 662-674.
Likert, R. 1961. New patterns ol management. New York: McGraw-Hill.
Mael, F., & Ashforth, B. E. 1992. Alumni and their alma mater: A partial test of the reformu-
lated model of organizational identification. Journal ol Organizational Behavior, 13:
103-123.
46 Academy oí Management Review January

Manz, C. C. 1986. Self-leadership: Toward an expanded theory of self-influence processes in

organizations. Academy oí Management Review, 11: 585-600.
Manz, C. C. 1990. Beyond self-managing work teams: Toward self-leading teams in the work-
place. In R. Woodman & W. Pasmore (Eds.), Research in organizational change and
development: 273-299. Greenwich, CT: JAI Press.
Maslow, A. H. 1970. Motivation and personality. New York: Harper & Row.
Mayer, R. C , & Schoorman, F. D. 1992. Predicting participation and production outcomes
through a two-dimensional model of organizational commitment. Academe/ oí Manage-
ment Journal 35:671-684.
Mayer, R. C , Davis, J. H., & Schoorman, F. D. 1995. An integrative model of organizational
trust. Academy oí Management Review, 20: 709-734.
McClelland, D. C. 1970. The two faces of power. Journal oí international Aííairs, 24: 29-47.
McClelland, D. C. 1975. Power: The inner experience. New York: Irvington
McClelland, D. C , & Burnham, D. H. 1978. Power is the great motivator. Harvard Business
Review. 54(2): 100-110.
McGregor, D. (1960). The human side oí the enterprise. New York: McGraw-Hill.
McGregor, D. 1966. Leadership and motivation. Cambridge, MA: MIT Press.
Molz, R. 1988. Managerial domination of boards of directors and financial performance.
Journal oí Business Research, 16: 235-249.
Mowday, R., Porter, L., & Steers, R. 1982. Organizational linkages: The psychology oí com-
mitment, absenteeism, and turnover. New York: Academic Press.
O'Reilly, C. 1989. Corporations, culture and commitment: Motivation and social control in
organizations. Caliiomia Management Review, 31(4): 9-25.
O'Reilly, C , & Chatman, J. 1986. Organizational commitment and psychological attachment:
The effects of compliance, identification, and internalization on prosocial behavior.
Journal oí Applied Psychology, 71: 492-499.
Perrow, C. 1986. Complex organizations: A critical essay. New York: McGraw-Hill.
Porter, L. W., Steers, R. M., Mowday, R. T., & Boulian, P. V. 1974. Organizational commitment,
job satisfaction, and turnover among psychiatric technicians. Journal oí Applied Psy-
chology, 5: 603-609.
Rechner, P. L., & Dalton, D. R. 1991. CEO duality and organizational performance: A longitu-
dinal analysis. Strategic Management Journal, 12: 155—160.
Salancik, G. R., & Meindl, J. R. 1984. Corporate attributions as strategic illusions of manage-
ment control. Administrative Science Quarterly, 29: 238-254.
Scandura, T. A., & Schriesheim, C. A. 1994. Leader-member exchange and supervisor career
mentoring as complementary constructs in leadership research. Academy oí Manage-
ment Journal, 37: 1588-1602.
Simon, H. A. 1957a. Administrative behavior (2nd ed.). Glencoe, IL: Free Press.
Simon, H. A. 1957b. Models oí man. New York: Wiley.
Simon, H. A. 1973. Organization man: Rational or self-actualizing? Public Administration
Review, 33 Quly/August): 346-353.
Smith, C. A., Organ, D., & Near, J. 1983. Organizational citizenship behavior: Its nature and
antecedents. Journal oí Applied Psychology, 68: 653-683.
Staw, B. M., McKechnie, P. I., & Puffer, S. M. 1983. The justification of organizational perfor-
mance. Administrative Science Quarterly. 28: 582-600.
1997 Davis, Schoorman, and Donaldson 47

Steers, R. M., & Black, J. S. 1994. Organizaíionaí Behavior. New York: HarperCollins.
Steers, R. M., & Porter, L. W. 1991. Motivation & work behavior. (5th ed.). New York: McGraw-
Hill.
Sussman, M., & Vecchio, R. P. 1982. A social influence interpretation of worker motivation.
Academy of Management Review, 7: 177-186.
Triandis, H. C. 1990. Cross-cultural studies of individualism and collectivism. In J. J. Berman
(Ed.), Nebraska symposium on motivation, vol. 37: 41—134. Lincoln: University of Nebraska
Press.
Triandis, H. C. 1995. ¡ndividualism and collectivism. Boulder, CO: Westview.
Triandis, H. C , Dunnette, M., & Hough, I. M. (Eds.). 1993. Cross-cultural studies. Handboot of
industrial and organizational psychology. Vol. 4. Palo Alto, CA: Consulting Psycholo-
gists.
Turner, J. C. 1981. The experimental social psychology of intergroup behavior. In J. C. Turner
& H. Giles (Eds.), Intergroup behavior: 66-101. Chicago: University of Chicago Press.
Walsh, J. P., Seward, J. K. 1990. On the efficiency of internal and external corporate control
mechanisms. Journal of Management Review, 15: 421-458.
Walton, R. E. 1980. Establishing and maintaining high-commitment work systems. In J. R.
Kimberly, R. H. Miles, & Associates (Eds.), The organizational life cycle: Issues in the
creation, transformation, and decline of organizations: 208-290. San Francisco: Jossey-
Bass.
Walton, R. E. 1985. From control to commitment in the workplace. Harvard Business Review,
63(2): 76-84.
Williamson, O. E. 1985. The economic institutions of capitalism: Firms, markets, relational
contracting. New York: Free Press.

James H. Davis is an assistant professor of corporate strategy and international

management at the University of Notre Dame. He received his Ph.D. in policy and
strategy from the University of Iowa. His research interests include stewardship
theory, trust, and corporate governance.
F. David Schoorman is an associate professor of organizational behavior and human
resource management at Purdue University. He received his Ph.D. in industrial ad-
ministraiton from Carnegie Mellon University. His interests include decision mak-
ing, motivation, trust, and work teams.
Lex Donaldson is a professor of organization design in the Australian Graduate
School of Management, in the University of New South Wales. He holds a Ph.D. from
the University of London. His research interest is theories of organization.
Copyright of Academy of Management Review is the property of Academy of Management and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.