March and Simon, Organizations, 1958

Ch. 1: Organizational Behavior

At the time of this book, little had been written on orgs. in the social sciences. M & S ask, is this is due to the unimportance of the subject ("no"), its distribution in other subdisciplines ("yes, to some extent"), or is it simply because little had been written ("yes")? Orgs are important because they are ubiquitous. Analogy to biology mentioned, "if we do not take it literally or too seriously"-- orgs, like animals, have highly specific structure, are highly coordinated.

A lot has been said by executives, managers, sociologists, social psychologists, political scientists, and economists, but these different "languages" have not been speaking to each other. They place emphasis on the need for a "common language" which they propose to do with their book. Also, there is great disparity between hypotheses and evidence in org studies- too much of former and not enough latter. They propose to remedy this as well.

Previous propositions about orgs can be generally grouped into three types:

1.) props that org members are passive instruments (employees or members who are given instructions but do not take action themselves) who behave in a stimulus-response mode,

2.) that org members bring attitudes, values, and goals to the org and must be induced to perform work, and

3.) that org members are decision makers and problem solvers. The authors argue that these 3 perspectives do not inherently "contradict" each other (they refer to them as the "instrumental," "motivational," and "rational perspectives"), though they are very different views on the same general subject.

The final section in this chapter is on psychology and rational action: actors have values about different outcomes (I prefer milk to orange juice), beliefs about relations between actions and outcomes (if I ask for milk, I will probably get milk and not orange juice), and multiple possible courses of actions (I can ask for milk verbally, I can write down a request, I can go to the store and buy the milk myself, etc.). "This, then is the general picture of the human organism that we will use to analyze organizational behavior. It is a picture of a choosing, decision-making, problem-solving organism [M&S constantly refer to people as "organisms"] that can do only one or a few things at a time, and that can attend to only a small part of the information recorded in its memory and presented by the environment. We shall see that these particular characteristics of the human organism are basic to some of the salient characteristics of human behavior in organizations."

Ch. 3: Organizational Constraints: Intraorganizational Decisions

Where chapter 2 (which is not in the reading list) focused on actor "instrumentality" (i.e., humans as simple machines or "instruments"); this chapter (chapter 3) is on "motivation." The "machine" model of chapter 2 ignores the range of roles that actors must perform- people, according to M&S, are more complicated than that. This chapter argues that individuals may have goals developed outside the org (from family, society, etc.) that may be in direct contradiction to org. goals. This potential source of conflict needs exploration.

Three models of bureaucracy are presented and examined in this chapter. I do not think that we should make any effort to memorize these three, or even achieve a thorough understanding of them, but we should probably know that they exist. I will give a brief sketch of each, but must admit here that my own understanding of them is limited, and encourage you to look at them in the original text if you really want a more detailed explanation than the one I am giving here.

The Merton model is concerned with dysfunctional organizational learning: individuals learn a certain set of stimulus-response techniques within the structure of an org, and then start applying this set to other contexts in which it is not appropriate (i.e., "dysfunctional"). An example: we learn to hear a ring, pick up a phone, and say "Hello." Having mastered this technique at home, we become able to apply it to a job as a receptionist. M&S argue that Personality is the connection between stimuli and given responses- the term "personality" means a response pattern that does not easily change or adapt. The demand for control by the org through the creation and enforcement of rules is an attempt to obtain a reliable response of org members to certain stimuli- to suppress personality as a factor. Continuing the example, suppose you were used to answering the phone by saying, "Hello, this is Connie," and now your company wants you to just say, "Wigits Incorporated." You may have some difficulty "altering" your old behavior to accommodate the new requirements of your job, and feel an urge to announce your name to the caller. This is personality, and orgs work to suppress it.

Where Merton emphasizes "rules" as a response for a demand for control, the Selznick model (as you will read in the review if the TVA in another article summary) emphasizes the delegation of authority: As those higher up in an org hierarchy delegate authority to those under them, there is both an increase in org goal achievement (more people doing more specialized tasks generally improves operating efficiency) and a simultaneous decrease in the same activity (multiple, increasingly powerful units now compete for resources and importance-- this is what happened to the TVA). Delegation results in internal struggles for legitimacy at the same time that it makes org members more committed to the organization.

The final model is the Gouldner model. It is the simplest of the three models, but has some of the same features as previous ones. like Merton, Gouldner is concerned with the consequences of bureaucratic rules for the "maintenance of organizational structure." Like both M. and S., G. he is concerned with how efforts to exert control in a particular part of a system (making everyone answer the phone the same way) can disturb the larger system (pretty soon everyone is answering the phone the same way, at home and at work) with feedback on the original subsystem (now, no one can tell if you are at work or at home when you answer the phone- this example is bizarre, but I hope it gives you the idea).

Taking this system-model to the next level, Gouldner argues that the feedback loops can become destructive: demand for control ("I want all the receptionists to answer the phones with a cheery voice and say 'Wigits Incorporated, how can I help you?'") leads to the creation of rules ("Okay, people, here is how you must answer the phone") which affects people's beliefs in what is the minimum acceptable behavior ("We receptionists can probably get away with answering the phones without being cheerful about it") which creates a need for closer supervision ("If I want the receptionists to answer the phones in the right way I have to stand over them and watch them do it") which in turn emphasizes the visibility of power relations (who has the right to monitor whom) that affects interpersonal tension (bosses and employees now hate each other) creating a need for greater rules. This circle is complicated- I don't think that understanding all of its complexity is necessary or particularly useful, but there it is, at least in part.

The point of all three models is this: people are more complex than simple machines that do what they are told. Employees make two types of decisions within organizations: the first is to participate or to leave; the second is to produce or refuse to produce as required. These decisions are separate and independent, and much of the confusion "in the literature" about motivation comes from confounding the two.

M&S go on to propose their own system for understanding motivation to produce. It, too, is complex and full of variables, but the idea remains the same: people are more complex than machines, and they bring their own "baggage" into organizations with them- sometimes they work with an organization to achieve its goals, and sometimes against it, for whatever reason. Orgs cannot create personalities, cannot mandate enthusiasm or performance, but can operate to facilitate or hinder these things to some degree or other.

Ch. 6 Cognitive Limits on Rationality

M&S begin this chapter with a re-iteration of their main tasks in the book. It is worth reading here again: to "eliminate, one by one, the artificialities of the classical description of the employee as an instrument, and to replace this abstraction with a new one that recognizes that members of organizations have wants, motives, and desires, and are limited in their knowledge and in their capacities to learn and to solve problems."

This chapter's emphasis, as you can guess from the title, is on rationality and the cognitive aspects of org. behavior. They compare the ideas of rational "administrative" man with classical models of rationality: the latter assume that, for all actors (read "workers" here) all alternatives of a particular action are "given," i.e., that all consequences of particular actions are "known" by the actors (which is the case in a world of perfect information and where actions produce knows results), and that actors have a "complete utility-ordering" (clearly defined set of preferences over outcomes). In the authors' model, choice is "always exercised with respect to a limited, approximate, simplified 'model' of the real situation" ("definition of the situation"), and the elements of the definition of the situation are not "given" - that is, "we do not take these as data of our theory- but are themselves the outcome of psychological and sociological processes."

That probably didn't make much sense. Let me try to explain: we, as actors, do not live in worlds of perfect information about the world we live in: sometimes a particular action does NOT produce the intended outcome, and sometimes we do not know what all of the possible actions are for achieving particular outcomes (there may be many ways to reach a given goal, whether we are aware of them all or not). Given this, M&S propose that actors act with the best tools available to them, rather than with all of the possible information that might or does exist in the world. In this way, knowledge and beliefs have an enormous influence on "rational" behavior.

Key term: Performance program- an organization's orchestrated and sometimes elaborate response to a particular environmental stimulus (when we hear a fire alarm, we leave a building). These stimulus-response behaviors account for a large part of human behavior, inside and outside orgs. Through observing behavior, interviewing org members, and examining org documents, performance programs can be discovered. Knowledge of these programs provides large predictive power. Basically, M&S argue that we often behave in ways we do not think about; we "know" what to do when a particular "thing" happens, and we respond, in Pavlovian fashion, to these stimuli.

Critical concepts: the difference between "optimizing" and "satisficing" as kinds of behavior. The former is an attempt to determine what is the best possible method to achieve the most desirable goal that can be conceived; the latter is an attempt to determine what method is sufficient to achieve the current necessary goals (see esp. p. 140, Orgs p. 130). Most human decision making is satisficing activity, NOT optimizing activity (according to our authors).

Orgs face a classical problem: in order to behave adaptively, they need stable regulations and procedures that they can use in carrying out their adaptive practices. Performance programs are part of the org structure, but are the least stable part (according to M&S)-- switching performance programs from one kind to another when the environment changes is done with relative ease: if the name of a company changes, getting the receptionists to state the new name of the company over the phone is a simple switch. Rules for switching from one program to another are more stable: the receptionists must be told when to start answering the phones in the new way, on a particular day after the name change is finalized.. Most stable still are the procedures for developing, elaborating, instituting, and revising performance programs: the determination of how receptionists should answer the phones, what they should say, how they should say it, and when they should change. These three exist in a hierarchy that is easy to visualize (no?).

And so they end the chapter with this quote: "The limits on the power, speed, and capacity of human cognitive abilities make this new perspective on rationality different from the classical models which suppose no limits." Actors are constrained by themselves- believe it or not, this was hot stuff back in 1958.

Ch. 7 Planning and Innovating in Organizations

The start to this chapter contains one of the more interesting assertions of the book: Orgs, like people, tend not to innovate when things are going smoothly; innovation is costly. "Where search for new alternatives is suppressed, program continuity is facilitated." In other words, when you are making your product and people are buying it and your employees are happy and you are making a profit and the shareholders are not complaining, there is no reason to invest in research an development to find "better" products or more efficient ways of producing the same product- M&S argue that people within orgs are motivated by external pressure, and that when pressure is absent, change is also absent.

The idea of sunk costs is brought up again: the development of a particular kind of industry or business represents a sunk cost in that you have made certain investments to get the industrial ball rolling (buying machinery, renting a warehouse, etc.); changing the direction or speed of the ball represents a waste (or at least a misuse) of those sunk costs, and will be avoided if possible. Hence, setting the ball rolling in a particular direction represents a "sunk cost of innovation" and will tend to promote "program continuity." (I.e., changing the direction of the ball (or "program") will tend to be avoided.)

Here is a summary of the essential characteristics of the rational choice planning model for understanding organizational change:

1. The main requirement of actors in an organization is to satisfy situational criteria, not optimize, with criteria possibly shifting over time.

2. When one or more criteria are not being met (e.g., sales are dropping, profit is disappearing, etc.), an action program will be initiated to remedy the situation (cut-backs in employee health insurance, drastic lay-offs, increased marketing, etc.-- some intentional plan for changing the current situation).

3. A change in an org program requires a process of initiation through which new program possibilities are generated and evaluated- an org that has never had to layoff employees before must figure out how to do this; an org that wants to recapture its market must plan a strategy for competing more effectively with its competitors.

4. Generally, action programs are associated with particular criteria- and the world is empty of complicated causal relations: if we do advertise more, we will probably have an increase in sales; if we layoff employees, we will have a short-term cut in overhead expenses. Action programs are related through the demands they make on organizational resources: a decision to layoff employees might provide the extra funds necessary to launch a new advertising campaign-- if a company has limited resources, this might be one effective way to re-establish a "satisfactory" economic situation for the organization.

Bounded rationality appears here again. It is always worth reiterating: "...the limits of information available to humans and their abilities to use information in their computations." Humans have realistic limits on planning capacity- cognition plays a large role in organizational change, from discovering problems to offering solutions to implementing them.

M&S end the chapter with a quick recap of their project and its history: the first quarter of the 20th century was devoted to the idea of humans as complicated machines; the second quarter to the idea that humans have feelings and values that influence their behavior; M&S are making the argument that people also can behave adaptively and rationally (within their God-given limits), and will attempt to do so when the situation calls for it. Their book is a very long-winded way of stating this, but that is its primary goal.