Th eories of Bureaucratic Politics 255
decisionmaking can be considered democratic. Although bureaucracy is well in-
sulated from the pressures of the democratic process, the ranks of the civil service
can represent a broad demographic cross-section of society. If varied enough, this
cross-section is enough to ensure that the major interests of the various groups
within society are included in bureaucratic decisionmaking. Th rough diversity
in the civil service ranks, bureaucratic policymaking can be considered repre-
sentative, and as such, can stake a legitimate claim to abiding by and upholding
democratic values.
Th e basic claims of representative bureaucracy have received considerable
empirical support and shown themselves to be capable of limited predictive as
well as descriptive and explanatory power (Selden 1997). Once again, however,
there are limits. It is not clear that some interests are embodied in demographics,
and at least some civil servants seem to be highly capable of representing interests
that are not associated with their demographic profi les. A further weakness of the
representative bureaucracy literature is the surprising lopsidedness of its empiri-
cal focus—there is a considerable body of literature examining the demographic
makeup of the bureaucracy, and a comparative paucity of studies examining the
link between this demographic variation and a given agency’s policy outputs.
Work by Nick Th eobald and Donald Haider-Markel (2009) has further demon-
strated that even when representative bureaucracy fails to produce tangible policy
benefi ts, shared demographic representation between citizens and bureaucrats
can produce attitudinal benefi ts.
Beginning with Waldo, the bureaucratic politics movement has thus far been
much more successful in demonstrating the need for political theories of bureau-
cracy than in actually creating comprehensive frameworks to fi ll that need. In the
scientifi c sense, then, theories of representative bureaucracy are still immature.
Th ey tend to be considerably less parsimonious and elegant than the positivist
ideal (a not-unusual characteristic of inductivism), and their focus on contextual
detail has presented diffi culties for replicability and predictive capacity.
Where bureaucratic politics theories shine is in their ability to make system-
atic sense of the oft en confusing arena of public administration as we fi nd it in
the real world. As a path to ordering the facts of administration coherently, the
bureaucratic politics movement has yielded important dividends. Such works as
Harold Seidman’s (1998), for example, represent an important leap in creating a
more realistic portrayal and understanding of administration. In practice, admin-
istration is not about effi ciency, or even eff ectiveness. It is about politics, and once
that basic fact is grasped, the confusing jumble of agencies and their roles and
relations to the rest of the polity becomes much easier to understand.
In the latter sense, theories of bureaucratic politics have served the discipline
well. In highlighting bureaucracy’s political role, they have forged a greater un-
derstanding of why public agencies do what they do. If there is one area in which
such theories may fall short, it is in elegance or parsimony of the models. As the
number of actors, institutions, and sectors involved in policymaking expands,