Managing by Contract 119
Managing by Contract
Th e contracting-out phenomenon goes by several names—government by proxy
(Kettl 1993b); third-party government (Smith and Lipsky 1993; Salamon 1989);
hollow government; the hollow state (Milward 1994, 1996; Milward and Provan
2000a); shadow government and the contracting regime (Kettl 1988; Egger 2006).
Contracting out is also a theory of the control of bureaucracy and of organization
theory. Here we will deal only with contemporary government contracting as a
theory of management.
Contracting out is a key feature of the contemporary doctrines of manage-
ment in public administration. Since the mid-1990s, a steadily increasing per-
centage of public activities has been carried out “indirectly” by contractors and
for virtually every conceivable government function (Kettl 1993b). Most theories
of management assume a contained or bounded institution with managerial re-
sponsibilities for directing the day-to-day internal functioning of the organi-
zation as well as responsibility for conducting boundary transactions that link
the organization to other organizations and to its publics. Th e work of public
administrators is increasingly not this kind of management; it is, instead, the
management of contracts. Virtually all the capital functions of state and local
governments have always been done by contract, primarily with architectural
fi rms; building contractors; bridge builders; highway construction and mainte-
nance fi rms; sewer, sanitation, water works, and systems companies; and dam
builders. Th e US Defense Department has always contracted for airplanes, ships,
tanks, guns, and war technology. Most of the work of the National Aeronautics
and Space Administration and the Department of Energy is done by contract. It
is estimated that one-sixth of total federal spending (including entitlement) has
gone to contractors (Milward 1994); this amount, as shown by federal spending
fi gures, has been more than matched during the 2000s. Th e federal government
has many more civilian employees on the “contract” payroll than on the actual
civilian payroll (Frederickson 1997b). It has been shown that cuts in the federal
workforce can result in an increase in the contract payroll (Light 1999).
With this background, we turn to issues of management theory, and ask these
questions: Can government by contract be properly considered a form of man-
agement? What would be the salient feature of a management theory that as-
sumed work is being done by contract?
Can government by contract be thought of as a form of public management?
Th e answer is yes. Contract management is, however, diff erent from traditional
management in many signifi cant ways and therefore requires a somewhat dif-
ferent theory. It is fashionable to use principal-agent theory to explain the con-
tracting regime as an organizational scheme, and the theory is useful for that
purpose. But principal-agent theory off ers little to management theory, particu-
larly when compared with traditional theory, which is based on the direct study
of managerial behavior and managerial options in contract circumstances. Th e