safely violate the oYcially posted speed limit because they know that police oYcers
are highly unlikely to stop them for driving slightly faster than the oYcial limit.
The ‘‘real’’ speed limit is determined by the enforcement decisions of the police.
Thus, to a powerful degree, bureaucratic power depends on decisions. Indeed,
Herbert Simon ( 1976 , 1 ; compare Barnard 1938 ) contended that ‘‘a theory of
administration should be concerned with the processes of decision as well as
with the processes of action.’’ Simon argued:
The task of ‘‘deciding’’ pervades the entire administrative organization quite as much as
does the task of ‘‘doing’’—indeed, it is integrally tied up with the latter. A general theory
of organization that will insure correct decision-making must include principles of
organization that will include correct decision-making, just as it must include principles
that will insure eVective action.
Understanding—and controlling—those decisions depends on information. But
that, in turn, helps identify bureaucratic pathologies. Formal theorists from within
economics agreed that information is essential and that information asymmetries
plague relationships within bureaucracies. They imagined bureaucracy as a series
of contracts between principals (the higher-level oYcial charged with responsibil-
ity for a policy) and agents (the lower-level oYcial charged with carrying it out).
Principals hire agents to do the bureaucracy’s job; agents agree to do it in return for
compensation. Such relationships cascade through bureaucracies, from top to
bottom. They tend to produce two pathologies. First, principals need to pick
good agents, but they can never know enough about the agent to make sure they
have made the right choices. Theorists call this ‘‘adverse selection,’’ and poorly
chosen agents might not have the ability—or the inclination—to carry out a policy
as the principal wants. Second, principals can never know enough about what the
agent does to ensure that the agent carries out the terms of the contract. Theorists
call this ‘‘moral hazard,’’ and the problem makes it hard to provide adequate
supervision: to detect and correct problems in getting the bureaucracy’s work
done (Coase 1937 ; Williamson 1975 ; Wood and Waterman 1991 ).
The task of making decisions, however, depends heavily on the bureaucrat’s
position within the bureaucracy. 1 As former US federal administrator Rufus Miles
( 1978 ) famously put it, ‘‘Where you stand depends on where you sit.’’ Indeed,
Michel Crozier’s powerful analysis, The Bureaucratic Phenomenon ( 1964 ),
concludes that bureaucratic institutions must be understood in terms of the
cultural context within which they work. Within bureaucracies, oYcials at diVerent
1 More generally, there is a rich tradition within political science of treating ‘‘bureaucracy’’ as a
political actor, and as an institution composed of political actors. In addition to Wilson 1989 , see
Allison 1969 , 1971 ; Derthick 1972 ; Halperin 1974 ; Pressman and Wildavsky 1974 ; Bardach 1977 ; and
Hogwood and Peters 1985. There is also an emerging tradition that traces the roots of bureaucratic
behavior to its historical development. See, in particular, Skowronek 1982 ; Carpenter 2001 ; and Orren
and Skowronek 2004.
public bureaucracies 369