76 4: Public Institutional Th eory
may be in public administration, Mintzberg’s famous elliptical parsing has be-
come the standard for the visual images of hierarchy and the language used to
describe those images. More importantly, in the private and public sectors these
visual images and this language form the basis of testable hypotheses having to do
with organizational structure and design.
Translating the categories of work from those commonly found in industry,
such as sales and marketing, to categories commonly found in the public sector,
such as legislative liaison and contract management, is relatively simple; and so
is the adaptation of Mintzberg’s model to the wide range of public sector insti-
tutions, namely, police departments, state departments of social services, the US
Department of Agriculture.
Th ompson argued that “uncertainty appears to be the fundamental problem
for complex organizations, and coping with uncertainty, is the essence of the
administrative process” (1967, 159). To protect itself from contextual buff eting,
the organization will tend to seal off its technical and operating core through the
standardizing of work processes (lots of rules), planning, stockpiling, professional
gatekeeping, training, the rationing of services, and so forth. Some organizations,
particularly in the public sector, dominate their environments because they are
the only legitimate source of service, the US Department of Defense being an
example.
If uncertainty is the dominant contextual problem for institutions, interde-
pendence is the primary internal problem. Among organization theorists, the
concept of coupling is most commonly used to explain patterns of interdepen-
dence. Tasks and fl ows of work can be coupled in three ways: sequentially, by
pooling, or reciprocally; and in all these, forms may be loosely or tightly coupled.
Perhaps the best illustrations of these concepts as they are applied in the pub-
lic sector are the extended series of research on high-reliability systems by Todd
R. LaPorte and his colleagues (LaPorte and Consolini 1991); Martin Landau’s
work (1991) on redundant systems; Cohen and March’s work (1986) on large
research universities as loosely coupled systems; March and Olsen’s work (1986)
on garbage cans; the H. Brinton Milward series (1996) on the hollow state and the
application of contract regimes; and Donald Chisholm’s consideration (1995) of
problem solving and organizational design.
Th e vast range and variety of hierarchies lend themselves to categories. Simple
hierarchical structures are associated with smaller and newer organizations that
emphasize direct supervision, centralization, and the strategic apex. Many non-
profi t organizations under contract to the public sector tend in the direction of
simple structures of this type. Machine bureaucracies tend to be older and larger
hierarchies in which work standardization is critical. In such hierarchies, as il-
lustrated by the US Postal Service, the technostructure is especially infl uential.
Professional bureaucracies standardize work according to skills rather than the
work, tend to be decentralized and loosely coupled, emphasize training and edu-
cation, and oft en deal with complex problems. In professional bureaucracies, the