Introduction: What Are Th eories of Control of Bureaucracy? 15
Dwight Waldo (1946) and Herbert Simon (1947/1997) challenged the dichot-
omy, each for diff erent reasons. To Waldo, all administrative acts were political
at a fundamental level. To Simon, it was diffi cult empirically to unbundle politics
from administration, and vice versa. So, from the 1950s through the 1970s, the
received wisdom was that there was no dichotomy. Th en in the 1980s, the dichot-
omy reemerged and is now alive and well and found in control-of-bureaucracy
theory.
Th e signifi cance of control-of-bureaucracy theory is that it provides for the
analysis of public administration by making distinctions between political and
administrative acts or actions and/or between political and administrative actors.
Th ese distinctions are especially useful analytically because they provide for the
parsing of variables on the basis of politics (usually independent variables) and
administration (usually dependent variables).
We come, then, to the second important assumption in the control-of-
bureaucracy theory: In democratic self-government, elected offi cials, including
legislators and executives (presidents, governors, mayors), should control the de-
cisions and actions of appointed (usually civil service) offi cials. In American po-
litical science, the form and character of political control over bureaucracy are a
long-standing debate about what ought to be the proper range of discretion given
bureaucracies and bureaucrats (Finer 1941; Friedrich 1940). In modern times, this
debate is best characterized, on the one hand, by the Th eodore Lowi (1979) argu-
ment that we need a juridical democracy in which laws and regulations are so pre-
cise and so limiting that they deny bureaucracies latitude in carrying out the law
and, on the other hand, the Charles Goodsell (1983) argument that a wide bureau-
cratic discretion is essential to achieve eff ective and humane fulfi llment of the law.
Donald Kettl captures these diff erences well and puts them in historic context:
Diff erent approaches to the study of administration usually come from one of
two confl icting traditions in American politics—and each tradition leads to a
very diff erent perspective on the role of administration in American democ-
racy. Some students of administration come to the subject with a fundamentally
Hamiltonian bent. Like Alexander Hamilton, they seek a vigorous state vested
with a strong administrative apparatus. Other students of administration, how-
ever, are fundamentally Madisonians. Like Madison, they see in a delicate bal-
ance of power the best protection against tyranny. Th e competition of political
interests, in their view, lessens the risk that bureaucracy can abuse individual
liberty. (1993a, 407)Th e control-of-bureaucracy theory draws deeply from the Madisonian well of
distrust of administrative power. Many control-of-bureaucracy theorists are from
those parts of American political science that are essentially Madisonian. Econo-
mists and theories of economics have colonized political science and tend also to