Th e Rational, Self-Maximizing Bureaucrat 203
center on claims that the underlying assumptions about individual behavior and
the institutional dysfunction deduced to follow from these premises, bear only
a passing relationship to the real world. In a wide-ranging examination of the
empirical foundations of the budget-maximizing bureaucrat, for example, Andre
Blais and Stephane Dion (1991) note that the evidence is mixed. Bureaucrats do
seem to request larger budgets, but it is not clear they profi t from them through
better salaries, increased reputations, or any of the other elements Niskanen was
trying to condense into an individual utility function. Although budget maximiz-
ing has become fi rmly entrenched as public administration folklore, it is not even
clear that bureaucrats pursue these strategies as a general pattern. Some stud-
ies have found evidence that bureaucrats routinely pursue minimizing strategies,
looking not for ever bigger budgets but “for solutions for problems within their
agency through mandate clarifi cation as well as organizational, planning, and in-
formation changes” (Campbell and Naulls 1991, 113). Francis Fukuyama’s (2014)
portrayal of how the US Forest Service’s failure to adapt to changes in the science
of forestry led to “political decay” shows that even when budget maximizing be-
havior is evident, effi ciency and sound public policy do not necessarily follow.
As Fukuyama writes, “its offi cials came to be more interested in protecting their
budgets and jobs than in the effi cient performance of their mission” (8).
Niskanen even acknowledged some of these problems, and suggested that
the budget-maximizing model developed in Bureaucracy and Representative
Government was incomplete. He added, “In an important sense, it was also
wrong!” (Niskanen 1994, 269). Th e empirical evidence since the 1960s, Ni-
skanen argued, suggests that what bureaucrats tend to maximize is their dis-
cretionary budgets rather than their overall budgets. Th e discretionary budget
is defi ned as “the diff erence between the total budget and the minimum cost
of producing the output expected by the political authorities” (274). Th is is a
subtle but important diff erence from the budget-maximization standpoint.
It means that bureaucrats are seeking to maximize control over their budgets
rather than the absolute size of their budgets. In addition to this nuanced shift
in the behavioral assumptions driving his model, Niskanen also suggested that
his original work seriously underestimated the role of the political sponsors in
monitoring bureaucracy, an issue having even more important implications for
the conclusions of Tullock and Downs. Th ese early rational choice theories of
bureaucracy tended to paint pictures of politically powerful bureaucracies that,
at least under certain conditions, could act almost unilaterally as bureaucratic
self-interest displaced their public missions. To use Tullock’s words, rather than
agents of implementation for democratic institutions, a form of “bureaucratic
free enterprise” could develop in which bureaucrats pursued their own goals.
Since then, empirical work has provided considerable evidence that bureaucra-
cies tend to be highly responsive both to their political principals and to public
opinion generally (Wood and Waterman 1994; see Chapter 2 in that book for
a more thorough discussion of this topic). Th is does not necessarily mean that