Organizational theory has moved a great distance from the beginning of the 20
th
century; it has
an even greater distance to go. The narrative of the development of organization theory begins
with scientific management (Taylor 1911),
1
continues through classical bureaucratic studies
(Weber and Gerth 1958) and so-called "natural"
2
approaches (Barnard 1938), but then splits into
two very different open-systems categories. One is a rational, economic/political branch, which
starts with bounded rationality (March and Simon 1958), moves through contingency theory
(Lawrence and Lorsch 1967), then (much) later blooms into the "new economics of
organization," based on theories that seek to explain the existence and structure of organizations
as solutions to cooperation or coordination problems; here the exemplar is transaction cost
(Williamson 1985) economics. The other is a natural, sociological/psychological branch, which
includes garbage can theory (March and Olsen 1986), population ecology (Hannan and Freeman
1977), and the new (sociological) insitutionalism (Meyer and Rowan 1977). How do these
approaches differ in terms of levels-of-analysis covered, explanatory power, and potential for
future growth? Following these two historical paths, two trends are clear: the sophistication and
combination of different levels of analysis continually improves, and with each additional
theoretical model, further puzzles are solved. However, with the deep chasm separating the
sociological/psychological and economic/political branches, future prospects look bleak for these
groups to engage in the type of constructive dialogue that could lead to further innovations.
The first few organizational scholars used relatively simple conceptual models to explain
organizational behavior, restricting their analyses to a single level of analysis and a few
assumptions. Taylor focused on workers on the shop floor (individual level of analysis). While
most of his work concerned prescriptions to improve productivity rather than grand theorizing, to
do so he introduced the concept of the rational worker, who performed his duties in a sub-
1
To prevent citation proliferation, all further citations will be to the same articles. In order to compress ninety years
into five pages, many authors have been skipped; these authors should be taken as examples, not a complete listing!
2
Here I follow Scott's typology(Scott 1998), categorizing approaches by two dichotomous variables. The first is
rational/natural, indicating whether the approach is principally rational (assuming goal-directed behavior, or a "logic
of consequences," at a minimum) or natural (assuming norm-directed behavior, or a "logic of appropriateness"). The
second is open/closed, indicating whether it treats organizations as open or closed systems; I find this to be
somewhat limiting, and will develop throughout a more complex level of analysis typology instead
