168 7: Decision Th eory
preoccupied with action, and particularly with unsubstantiated “principles” of
action, which he labeled “proverbs.” Th e modern study of administration needed
to be more scientifi c, and the science of administration needed to be based on
a new and diff erent unit of analysis—the decision. Th e scientifi c focus on this
unit of analysis is still a hallmark of the conceptual framework theory, and con-
temporary scholarship is currently combining advances in several fi elds into a
rapid new evolution of decision theory. Whereas in the late 1990s the edge of
decision theory in public administration was defi ned by the tenets of bounded ra-
tionality, recent advances in psychology, sociology, and the relatively new fi eld of
behavioral economics have created a new concept of decisionmaking that might
be described as “predictable irrationality.”^1 In this newer framework, decisions
are seen not simply as a product of humans whose rationality is bounded by cog-
nitive and environmental limits, but also as a product of evolutionarily designed
and predictable patterns of cognitive information processing. We discuss these
advances later in the chapter.
Epistemologically, the fi rm scientifi c grounding of decision theory is hardly
surprising given its intellectual origins. Certainly as conceived by Simon, a theory
of decisionmaking has to be based on the logical positivist argument that there
must be a primary distinction between facts, which can be tested and verifi ed, and
individual and collective preferences and values, which cannot be scientifi cally
verifi ed. At the base of Simon’s decision theory is the concept of effi cient admin-
istrative rationality: “Th e correctness of an administrative decision is a relative
matter—it is correct if it selects appropriate means to reach designated ends. Th e
rational administrator is concerned with the selection of these eff ective means”
(1947/1997, 72). Rationality is based on means-ends logic, and assumes that the
question at hand is selecting the best means to achieve agreed-upon ends. Facts
and values “are related to means and ends. In the decision-making process those
alternatives are chosen which are considered to be appropriate means for reach-
ing desired ends. Ends themselves, however, are oft en merely instrumental to
more fi nal objectives” (61). We are thus led to Simon’s conception of a series, or
hierarchy, of ends: “Rationality has to do with the construction of means-ends
chains of this kind” (62). He noted the limitations to the logic of means-ends
analysis and the hierarchy of ends: Because means and ends can never be entirely
separated, ends are oft en incomplete and unclear, and both means and ends are
infl uenced by time and changing circumstances.
Rationality is therefore limited (later called bounded or limited). Nevertheless,
the science of administration should be based on the analysis of decisions, the
decision being the proper unit of analysis for our research and theory. “In the
fi rst place, an administrative science, like any science, is concerned purely with
factual statements. Th ere is no place for ethical assertions in the body of science.
Whenever ethical statements do occur, they can be separated into two parts, one
factual and one ethical; and only the former has any relevance to science” (Simon
1947/1997, 253).