120 5: Th eories of Public Management
“fi ndings” based on these observations would include those discussed in the
following paragraphs.
First, the oldest and still one of the largest forms of contracting out is for cap-
ital construction and repair and maintenance purposes, particularly at the state
and local levels of government. To be eff ective, such contracts must be highly pre-
cise regarding the quality of materials and workmanship, schedule, and the like.
Contractors are ordinarily held to very tight specifi cations. Th e contracting gov-
ernment, particularly for large-scale contracts, ordinarily has an on-site project
manager who oft en has czarlike powers and who carefully watches to see that the
contractor complies with specifi cations. Th e contracting fi rm is usually organized
and managed in the so-called project management style, which brings together
several skills and technologies. Large contracting fi rms almost always use a ma-
trix organizational form showing established skill-based departments that assign
workers to projects (Davis and Laurence 1977). Th e critical point in the contract
is the link between the government’s project manager (principal) and the fi rm’s
project manager (agent). To be eff ective, a combination of extensive technologi-
cal knowledge, trust, and toughness must be evident (Donahue 1989).
One of the best studies of contracting and privatization concludes that con-
tracting works best when (1) what needs to be done can be clearly and precisely
described for purposes of contract negotiation and compliance; (2) desired out-
comes can readily and easily be measured or identifi ed; (3) penalties are imposed
for noncompliance with the contract; and (4) contractors may be discontinued or
changed (Donahue 1989). Contracting for the construction and maintenance of
buildings, highways, bridges, and other capital facilities tends to meet these cri-
teria. As long as the four criteria are met, principal-agent theory can adequately
explain management by contract. Th e reason for our skepticism about the value
of principal-agent theory here is that these criteria are rarely met.
Th e key theoretical points here are the diff erences between a continuing per-
manent hierarchy and a contract. In the former, motivation, worker acceptance
of directives, group behavior, role diff erentiation, and managerial style are all crit-
ically important. In the latter, these requirements are exported to the contracting
fi rm, leaving government management to entail contract monitoring and over-
sight. In capital contracting for discrete projects, there are usually many qualifi ed
bidders, or, put another way, a genuine market and a capacity on the part of the
government to get the best product at the best price (Rehfuss 1989).
Because of cost controls on government, fewer and fewer public employees
become responsible for more and more contracts. Th is situation reduces contract
management to paper shuffl ing and auditing (Kettl 1993b; Rehfuss 1989; Cigler
1990; DeHoog 1984). State and local governments have diminished the capacity
for contract management and oversight and are gradually being hollowed out.
Th e problem, then, is not the theory of management by contract so much as it is
the application of the New Public Management concepts, and particularly con-
tracting out, without providing for eff ective contract management.