Pfeffer, J., Organizations and Organization Theory, Pittman, 1982, 1-23.


Summary

This paper is the introductory chapter to Pfeffer's book surveying organizational theories. His approach is "less unifying yet more comprehensive" than Scott's text. He is most concerned with the lack of focused effort on delineating and furthering organizational theory, and the over focus on theory generation and application in areas like leadership and job satisfaction.

Pfeffer chooses to organize existing theories around three "perspectives of action".

1. Action as Purposive, Bounded, Intended, Goal-Directed

This perspective closely follows Scott's rational system definition. where individuals act based on rational decisions against specific goals. Pfeffer's main critique with this perspective is that given the complexity of causal effects on behavior it's possible to find rationality after-the-fact in almost any action.

2. Action as Externally Constrained and Situational

This perspective is roughly part of Scott's natural system group. Individual action is determined by the social and environmental constraints acting on them.

3. Action as Random and Emergent

This perspective (part of Scott's natural system focus) is based on Weick's view that action is largely "random and unfolding". Action "happens" and later people apply either individualistic or structuralistic viewpoints to explain behavior. Being random, action is impossible to predict.

In addition to methodological differences, each viewpoint has contrasting views on the role of the manager. While rationalists see the manager as instrumental for organizational adaptation, the other perspectives see managers as either constrained by their environment or merely "figureheads" trying to maintain the illusion of rationality.

Levels of Analysis

In his introduction, Pfeffer sees two levels of analysis -- the individual/sub-unit level and the organizational level. He criticizes the use of aggregate data from individuals to explain organizational behavior, as well as using organizational or aggregate data to predict individual action.


Notes

This is the introduction of Pfeffer's book on organizational theory. His attempt is to be somewhat more comprehensive that Scott's book, at the expense of being less systematic and unifying. Pfeffer feels that there hasn't been effective delineation of organizational theory, and the existing theories are blurred by extensive focus on application and practice, and by overfocus on such topics as leadership and job satisfaction (at the expense of developing and testing theories).

Pfeffer organizes his views on theories into the following perspectives:

Perspectives on Action

Pfeffer differentiates theories on how they consider action in the organization. The three perspectives are:

* action as purposive, boundedly or intendedly rational, goal directed

* action as externally constrained or situationally determined

* action as random and emergent from social processes

Rational in #2 and #3 are constructed "after the fact" to account for behaviors. Other authors (Van de Ven) characterize action as "deterministic vs voluntaristic".

Prospective, Intendedly Rational, Created Action

The perspective of action as rational and foresightful has dominated theories of organizations. According to these views, choice precedes action and is directed toward goals. One problem with this view is that on can usually, after the fact, develop a set of choice preferences and rationality to explain almost any action.

To counter the economic "rational man" model that asssumes perfect knowledge and reasoning, Simon advanced the idea of "bounded rationality" to account for cognitive limits of individuals. Rather than optimizing, people take the first action that "saticfices" the objectives. Decision-making is thus based on experience and heuristics.

Various theories expound different interpretations and versions of bounded rationality -- some focus on developed rules, others acknowledge some of the emergent randomness of action.

 

Some of the recent theories more or less based on rationality are:

Individual level

* expectancy theory (Edwards)

* path-goal theory of leadership (House)

* goal-setting theory (Locke)

* needs theory and task design theory

Organizational Level

* structural contingency theory

* market failures or tranactions cost approach (Williamson)

* Marxist approaches

 

External Constraint or Situational Control Perspective

In this perspective actors are constrained and even determined by external forces and situation. Action results from the patterns of constraints and demands on the organization.

In this case rationality is used to make sense of the actions that have already occured in response to the environment. It explains action using variables from the environment and not entity-specific things like values, needs, and personalities. Some theories in this realm are:

* population ecology (Hannan and Freeman)

* operant conditioning (Luthans and Kreitner)

Other theories are less restrictive and allow some action by choice such as

* role theory

* social influence and social information processing

 

Almost Random, Emergent Process View of Action

This perspective is very different from the above two views. It sees action as neither rational nor deterministic, but random and unfolding. Things happen, and then rationality and understanding of the environment constraints is used to explain "history". It is impossible to predict what will happen even if you understood all the actors and intentions.

 

Some of the theories in this realm are:

* decision process theories (Cohen, March, Olsen)

* cognitive perspectives on organizations (Weick) sees system of shared meaning

 

Differences in Practice and Methodology

Each viewpoint has different ways of understanding and predicting (or not predicting behavior). The rational viewpoint assume administrative tasks improve performance, and have developed many refined tools for analysis and simulation. The situational viewpoint minimizes the affective role of the manager, and often uses case studies and simulations for evaluation. The random perspective inherantly doubts any success af prediction and sees the manger as a figurehead trying to maintain the semblance of rationality.

The methodological differences somewhat separate the theories as well, though portions of the perspectives can be combined with further analysis.

Levels and Units of Analysis

Pfeffer uses to categories of analysis -- individuals and subunits for one and the organization as a unit as the other.

There are methodological implications for the choice of analysis level. If one misses key variables in model development, aggretating to a larger social unit will overinflate the relationships between variables. Also extending conclusions made at the ecological level to individual behavior can also lead to errors. "The unit of analysis should correspond to the level of the theoretical mechanisms that are presumed to be affecting the dependent variables."

However, choosing the correct level of analysis can be problematic, especially when one can aggregate data to almost any level. Yet the poor choice of analysis level can lead to inaccurate conclusions. If you assume subunits as independent in an environment where they are collectively influences at a higher level, you may miss some important causal variables.

Most organizational studies have used the individual as the unit of analysis. Thus focus is on individual level variables (demographics, needs, attitudes), at the exclusion of normative and technological variables that operate collectively.

 

The Individualist-Structuralist Controversy

One one hand, some researchers like Weick reject the structural view of organizations and see behavior as a collection of micro-situations of individual behavior and relationships that evolve over time. They purport to evaluate social-level phenomena by studying the interlocked cycles of individual behavior. Organizations are "only abstractions from the behavior of individuals in time and space". While aggregate summaries of individual behavior are important, they are insufficient to empirically explain social processes.

The structuralist position believes that one can understand organizations without studying the microprocesses. There are enduring realities of things like role, formalization, centralization, and organizational structure. Collectivities are more than the sum of their parts. They belive its a fallacy to assume one can understand sociological behavior by studying only individuals. Individual analysis begins a reductionist approach that will eventually end with biological analyses. They also feel that individualists rely on hypothetical constructs of "society" that exist in people's heads.

Structuralists are concerned with aggregate properties of populations and the emergent properties of organizations themselves. Much less analysis has been done at this level than at the individual level.