chief executive's long service in a

chief executive's long service in an industry and his
or her hesitance to diversify from that industry.
But, in general, the perspective proposed here has
not been put to systematic or comprehensive test.
One reason may be that inquiry into the linkages
among individuals, organizations, and their competitive
environments necessarily requires a multidisciplinary
approach. A gulf, however, continues to
separate psychologists, sociologists, and researchers
with a strategy or economic orientation. It would be
the rare researcher who could draw equally on all
camps. The present writers recognize their own limitations
in this respect: this paper takes a lopsidedly
macro view while making relatively crude assumptions
about the psychological processes of top managers.
It is hoped that future research on the topic
will draw these disciplines together, allowing each
to build on the others.
Inquiry into the upper echelons perspective may
provide three major benefits. For the scholar, it
may ofl"er substantially greater power to predict organizational
outcomes than current theories afl'ord.
A second benefit may come to those responsible for
selecting and developing upper level executives. For
example, light may be shed on the tendencies of organizations
led by older executives, those with formal
management education, or those whose dominant
career emphasis has been in a particular
functional area. The effect of, say, management
teams with long term, stable membership, as opposed
to teams with short lived membership, also
may become more apparent. A third benefit may
accrue to the strategist who is trying to predict a
193
competitor's moves and countermoves. Can it be
demonstrated, for example, that a competing firm
headed by a team of executives who rose primarily
through operations will tend to be sluggish in responding
to a new product initiative? Or that a
chief executive brought in from outside the industry
will tend to steer the firm into new businesses, thus
making the core business relatively vulnerable in
the short run?
This paper has three primary aims. The first is to
propose a model of how upper echelon characteristics
may become reflected in organizational outcomes.
The second is to review literature that has
addressed the upper echelons perspective. The third
is to provide a foundation and stimulus for empirical
research into the links between managerial
backgrounds and organizational outcomes. To meet
this third aim, the paper identifies some major variables
of interest, propositions, and methodological
suggestions.
Development of the Model
Reconciliation with the Inertial Perspective
The view taken here is that top executives matter.
The contrary view—that large organizations
are swept along by events or somehow run themselves—
has been argued directly by Hall (1977)
and indirectly by the population ecologists (Hannan
& Freeman, 1977).
The most commonly cited empirical evidence
of the inertial organization is Lieberson and
O'Connor's (1972) study of top executives in large
corporations. Although an important study, it falls
short of being a definitive test of the impact of different
types of chief executives. First, it sought to
determine the different impacts of successive chief
executives within firms. Because new chief executives
of large firms predominantly are promoted
from within the firm and often are even "groomed"
by the outgoing chief executive, it is not surprising
that the authors found blurs between such eras. A
research design that highlights differences across
organizations would be a fairer test of whether different
types of managers are associated with different
organizational outcomes. Second, the Lieberson
and O'Connor study employed a combination of dependent
variables and data analysis that made it almost
impossible for the leadership variable to take
a major role. Two of their three dependent vari-
194
ables—dollar sales and earnings—are primarily indicators
of the firm's size and the type of industry it
is in. The third variable—return on sales—is closer
to being a universal performance indicator, but it,
too, carries a large industry-specific component and
so is not as good a measure as return on investment
or, even better, return on investment relative to the
industry. In their data analysis, the authors sought
first to explain variance in their performance measures
by using three independent variables: year, industry,
and company. Then the analysis was rerun
with leadership—a set of dummy variables—included
to determine how much additional variance
could be explained. As might be expected, the first
three independent variables were potent predictors
(as high as .97) of the performance measures, so
the apparent added effect of leadership was nil.
Thus, Lieberson and O'Connor's approach, which
also was used by Salancik and Pfeffer (1977) in
their study of the effect of mayors on city budgets,
is not an appropriate test: (1) it does not allow leadership
to enter earlier into the equation, and (2) the
equation is almost tautological given the choice
of independent and dependent variables. Weiner
and Mahoney (1981) attempted to overcome
these problems in a replication of Lieberson and
O'Connor's study and found that their "stewardship"
variable accounted for 44 percent of the variance
in profitability of major firms. The point here
is not to denigrate earlier research, but rather to
note the methodological complexities in such studies
and to observe that definitive findings on the unimportance
of chief executives are not in hand.
Human Limits on Choice
Theorists of the Carnegie School have argued
that complex decisions are largely the outcome of
behavioral factors rather than a mechanical quest
for economic optimization (Cyert &. March, 1963;
March & Simon, 1958). In their view, bounded rationality,
multiple and conflicting goals, myriad options,
and varying aspiration levels all serve to limit
the extent to which complex decisions can be made
on a techno-economic basis. Generally, the more
complex the decision, the more applicable this behavioral
theory is thought to be. So, for that class
of choices called "strategic"—complex and of major
significance to the organization—the behavioral
theory is especially apt.
The term "strategic choice" is used here in the