Basic Concepts in the Behavioral Theory of the Firm: Four major relational concepts

In the course of developing the three subtheories, we have developed a relatively small number of relational concepts. In many respects, they represent the heart of our theory of business decision making. The four major concepts used in the theory are (1) quasi resolution of conflict, (2) uncertainty avoidance, (3) problemistic search, and (4) organizational learning. In this section we review briefly the meaning of each of these concepts. In subsequent chapters we will use the concepts to build a number of models and to suggest implications for economic and organizational theory.

1. Quasi resolution of conflict

In keeping with virtually all theories of organizations, we assume that the coalition represented in an organization is a coalition of members having different goals. We require some procedure for resolving such conflict. The classic solution is to posit an exchange of money from some members of the coalition to other members as a way of inducing conformity to a single, consistent set of goals — the organizational objective.

We propose an alternate concept of organizational goals and an alternate set of assumptions about how conflict is resolved. Basically we have argued that most organizations most of the time exist and thrive with considerable latent conflict of goals. Except at the level of nonoperational objectives, there is no  internal consensus. The procedures for “resolving” such conflict do not reduce all goals to a common dimension or even make them obviously internally consistent.

Goals as independent constraints. In our framework,organizational goals are a series of independent aspiration-level constraints imposed on the organization by the members of the organizational coalition. These constraints may include nonessential demands (i.e., demands that are already satisfied when other constraints are met), sporadic demands (i.e., demands that are made only occasionally), nonoperational demands (i.e., demands for which there are no operational measures), as well as essential, continuous, operative goals. In general, although we recognize the importance of goals that are nonessential (because they might become essential), of goals that are ordinarily sporadic (because they occasionally are enforced), and of goals that are nonoperational (because they sometimes can be made operational), we will focus on those constraints that are essential, continuous, and operative.

Specifically, in the case of price and output models of the business firm, we assume a profit goal, a sales goal, a market share goal, an inventory goal, and a production goal. In any particular firm we expect some subset of these objectives to  be essential, continuous, and operative. Moreover, we expect that subset to pose problems for the organization in the form of potential conflict. Thus, we require assumptions about procedures for resolving conflict. We assume that conflict is resolved by using local rationality, acceptable-level decision rules, and sequential attention to goals.

Local rationality. We assume that an organization factors its decision problems into subproblems and assigns the subproblems to subunits in the organization. From the point of view of organizational conflict, the importance of such local rationality is in the tendency for the individual subunits to deal with a limited set of problems and a limited set of goals. At the limit, this reduces to solving one problem in terms of only one goal.

The sales department is primarily responsible for sales goals and sales strategy; the production department is primarily responsible for production goals and production procedures; the pricing department is primarily responsible for profit goals and price decisions; and so on.

Through delegation and specialization in decisions and goals, the  organization reduces a situation involving a complex set of interrelated problems and conflicting goals to a number of simple problems. Whether such a system will in fact “resolve” the conflict depends, of course, on whether the decisions generated by the system are consistent with each other and with the demands of the external environment. In our theory consistency is facilitated by two characteristics of the decision process: (1) acceptablelevel decision rules; (2) sequential attention to goals.

Acceptable-level decision rules. In the classic arguments for decentralization of decision making, we require strong assumptions about the effectiveness of the “invisible hand” in enforcing proper decisions on a system of local rationality. Consistency requires that local optimization by a series of independent decision centers result in over-all optimization. On the other hand, we are persuaded that organizations can and do operate with much weaker rules of consistency (i.e., we require that local decisions satisfying local demands made by a series of independent decision centers result in a joint solution that satisfies all demands). Such rules are weaker in two senses: (1) There will ordinarily be a large number of local decisions that are consistent with other local decisions under such a rule. The demand constraints do not uniquely define a solution. (2) Any such system will tend to underexploit the environment and thus leave excess resources to absorb potential inconsistencies in the local decisions.

Sequential attention to goals. Ordinarily when we talk of “consistency” of goals or decisions we refer to some way of assessing their internal logic at a point in time. As a result, in classic theories of organizations we are inclined to insist on some consistency within a cross section of goals. Such an insistence seems to  us inaccurate as a characterization of organizational behavior. Organizations resolve conflict among goals, in part, by attending to different goals at different times. Just as the political organization is likely to resolve conflicting pressures to “go left” and “go right” by first doing one and then the other, the business firm is likely to resolve conflicting pressures to “smooth production” and “satisfy customers” by first doing one and then the other. The resulting time buffer between goals permits the organization to solve one problem at a time, attending to one goal at a time.

2. Uncertainty avoidance

To all appearances, at least, uncertainty is a feature  of organizational decision making with which organizations must live. In the case of the business firm, there are uncertainties with respect to the behavior of the market, the deliveries of suppliers, the attitudes of shareholders, the behavior of competitors, the future actions of governmental agencies, and so on. As a result, much of modern decision theory has been concerned with the problems of decision making under risk and uncertainty. The solutions involved have been largely procedures for finding certainty equivalents (e.g., expected value) or introducing rules for living with the uncertainties (e.g., game theory).

Our studies indicate quite a different strategy on the part of organizations. Organizations avoid uncertainty: (1) They avoid the requirement that they correctly anticipate events in the distant future by using decision rules emphasizing short-run reaction to short-run feedback rather than anticipation of long-run uncertain events. They solve pressing problems rather than develop long-run strategies. (2) They avoid the requirement that they anticipate future reactions of other parts of their environment by arranging a negotiated environment. They impose plans, standard operating procedures, industry tradition, and uncertainty-absorbing contracts on that environment. In short, they achieve a reasonably manageable decision situation by avoiding planning where plans depend on predictions of uncertain future events and by emphasizing planning where the plans can be made self-confirming through some control device.

Feedback-react decision procedures. We assume that organizations make decisions by solving a series of problems; each problem is solved as it arises; the organization then waits for another problem to appear. Where decisions within the firm do not naturally fall into such a sequence, they are modified so that they will.

Consider, for example, the production-level decision. In most models of output determination, we introduce expectations with respect to future sales and relate output to such predictions. Our studies indicate, to the contrary, that organizations use only gross expectations about future sales in the output decision. They may, and frequently do, forecast sales and develop some long-run production plans on paper, but the actual production decisions are more frequently dominated by day-to-day and week-to-week feedback data from inventory, recent sales, and salesmen.

This assumption of a “fire department” organization is one of the most conspicuous features of our models. Under a rather broad class of situations, such behavior is rational for an organization having the goal structure we have postulated. Under an even broader set of situations, it is likely to be the pattern of behavior that is learned by an organization dealing with an uncertain world and quasi-resolved goals. It will be learned because by and large it will permit the organization to meet the demands of the members of the coalition.

Negotiated environment. Classical models of oligopoly ordinarily assume  that firms make some predictions about the behavior of their environment, especially those parts of the environment represented by competitors, suppliers, customers, and other parts of the organization. Certainly such considerations are important to any decisions made by the firm. Our studies, however, lead us to the proposition that firms will devise and negotiate an environment so as to eliminate the uncertainty. Rather than treat the environment as exogenous and to be predicted, they seek ways to make it controllable.

In the case of competitors, one of the conspicuous means of control is through the establishment of industry-wide conventional practices. If “good business practice” is standardized (through trade associations, journals, word of mouth, external consultants, etc.), we can be reasonably confident that all competitors will follow it. We do not mean  to imply that firms necessarily enter into collusive agreements in the legal sense; our impression is that ordinarily they do not, but they need not do so to achieve the same objective of stability in competitive practices.

For example, prices are frequently set on the basis of conventional practice. With time, such variables as the rate of mark-up, price lines, and standard costing procedures become customary within an industry. Some effects of such practices are indicated in Chapter 7. The net result of such activity with respect to prices (and comparable activity with regard to suppliers and customers) is that an uncertain environment is made quite highly predictable.

Such negotiation among firms is not obviously collusion for profit maximization. Rather, it is an attempt to avoid uncertainty while obtaining a  return that satisfies the profit and other demands of the coalition. The lack of a profit-maximizing rationale is suggested by (1) the stability of the practices over time and (2) the occasional instances of success by firms willing to violate the  conventional procedures (e.g., discount houses in retailing).

In a similar fashion, the internal planning process (e.g., the budget) provides a negotiated internal environment. A plan within the firm  is a series of contracts among the subunits in the firm. As in the case of industry conventions, internal conventions are hyperstable during the contract period and tend to be relatively stable from one period to the next (e.g., in resource allocation). As a result, they permit each unit to avoid uncertainty about other units in making decisions.

3. Problemistic search

In the framework proposed in this volume, the theory of choice and the theory of search are closely intertwined. Necessarily, if we argue that organizations use acceptable-level goals and select the first alternative they see that meets those goals, we must provide a theory of organizational search to supplement the concepts of decision making. In our models we assume that search, like decision making, is problem-directed. By problemistic search we mean search that is stimulated by a problem (usually a rather specific one) and is directed toward finding a solution to that problem. In a general way, problemistic search can be distinguished from both random curiosity and the search for understanding. It is distinguished from the former because it has a goal, from the latter because it is interested in understanding only insofar as such understanding contributes to control. Problemistic search is engineering rather than pure science.With respect to organizational search, we assume three things:

  1. Search is motivated. Whether the motivation exists on the buyer or seller side of the alternative market, problemistic search is stimulated by a problem, depressed by a problem solution.
  2. Search is simple-minded. It proceeds on the basis of a simple model of causality until driven to a more complex one.
  3. Search is The way in which the environment is viewed and the communications about the environment that are processed through the organization reflect variations in training, experience, and goals of the participants in the organization.

Motivated search. Search within the firm is problem-oriented. A problem is recognized when the organization either fails to satisfy one or more of its goals or when such a failure can be anticipated in the immediate future. So long as the problem is not solved, search will continue. The problem is solved either by discovering an alternative that satisfies the goals or by revising the goals to levels that make an available alternative acceptable. Solutions are also motivated to search for problems. Pet projects (e.g., cost savings in someone else’s department, expansion in our own department) look for crises (e.g., failure to achieve the profit goal, innovation by a competitor). In the theory we assume that variations in search activity (and search productivity) reflect primarily the extent to which motivation for search exists. Thus, we assume that regular, planned search is  relatively unimportant in inducing changes in existing solutions that are viewed as adequate.

Simple-minded search. We assume that rules for search are simpleminded in the sense that they reflect simple concepts of causality. Subject to learning (see below), search is based initially on two simple rules: (1) search in the neighborhood of the problem symptom and (2) search in the neighborhood of the current alternative. These two rules reflect different dimensions of the basic causal notions that a cause will be found “near” its effect and that a new solution will be found “near” an old one.

The neighborhood of symptom rule can be related to the subunits of the organization and their association with particular goals and with each other. A problem symptom will normally be failure on some goal indicator. Initial reaction, we assume, will be in the department identified with the goal. Thus, if the problem is the failure to attain the sales goal, the search begins in the sales department and with the sales program. Failing there, it might reasonably proceed to the problem of price and product quality and then to production costs.

The neighborhood of existing policy rule inhibits the movement of the organization to radically new alternatives (except under circumstances of considerable search pressure). Such an inhibition may be explained either in terms of some underlying organizational assumptions of continuity in performance functions or in terms of the problems of conceiving the adjustments required by radical shifts.

When search, using the simple causal rules, is not immediately successful, we assume two developments. First, the organization uses increasingly complex (“distant”) search; second, the organization introduces a third search rule:  (3) search in organizationally vulnerable areas.

The motivation to search in vulnerable areas stems from two things. On the one hand, the existence of organizational slack will tend to lead search activity in the direction of slack parts of the organization. On the other hand, certain activities in the organization are more easily attacked than others, simply because of their power position in the system. One general phenomenon is the vulnerability of those activities in the organization for which the connection with major goals is difficult to calculate concretely (e.g., research in many firms). In either case, a solution consists in either absorbing slack or renegotiating the basic coalition agreement to the disadvantage of the weaker members of the coalition.

Bias in search. We assume three different kinds of search bias: (1) bias reflecting special training or experience of various parts of the organization, (2) bias reflecting the interaction of hopes and expectations, and (3) communication biases reflecting unresolved conflict within the organization. Bias from prior experience or training is implicit in our assumptions of search learning (below), local specialization in problem solving (above), and subunit goal differentiation (above). Those parts of the organization responsible for the search activities will not necessarily see in the environment what those parts of the organization using the information would see if they executed the search themselves. The bias in adjusting expectations to hopes has the consequence of decreasing the amount of problem-solving time required to solve a problem and of stimulating the growth of organizational slack during good times and eliminating it during bad. We assume that communication bias can be substantially ignored in our models except under conditions where the internal biases in the firm are all (or substantially all) in the same direction or where biases in one direction are located in parts of the organization with an extremely favorable balance of power.

4. Organizational learning

Organizations learn: to assume that organizations go through the same processes of learning as do individual human beings seems unnecessarily naive, but organizations exhibit (as do other social institutions) adaptive behavior over time. Just as adaptations at the individual level depend upon phenomena of the  human physiology, organizational adaptation uses individual members of the organization as instruments. However, we believe it is possible to deal with adaptation at the aggregate level of the organization, in the same sense and for the same reasons that it is possible to deal with the concept of organizational decision making.

We focus on adaptation with respect to three different phases of the  decision process: adaptation of goals, adaptation in attention rules, and adaptation in search rules. We assume that organizations change their goals, shift their attention, and revise their procedures for search as a function of their experience.

Adaptation of goals. The goals with which we deal are in the form of aspiration levels, or — in the more general case — search equivalence classes. In simple terms, this means that on each dimension of organizational goals there are a number of critical values — critical, that is, from the point of view of shifts in search strategy. These values change over time in reaction to experience, either actual or vicarious.

We assume, therefore, that organizational goals in a particular time period are a function of (1) organizational goals of the previous time period, (2) organizational experience with respect to that goal in the previous period, and (3) experience of comparable organizations with respect to the goal dimension in the previous time period. Initially at least, we would assume a simple linear function,

where G is the organizational goal, E the experience of the organization, C a summary of the experience of comparable organizations, and where

. The parameters in this goal adaptation function are important attributes of the organization. a 3 reflects the organization’s sensitivity to the performance of competitors or other comparable organizations. a 1 and a 2 reflect the speed at which the organization revises goals in the face of experience. In some cases, we will want to define two values for a 3 — one for when comparative experience exceeds the organization’s goal and a different one for when it is below the goal. Similarly, we may want to allow the effect of the organization’s experience to depend on whether it exceeds or is below the goal.

Adaptation in attention rules. Just as organizations learn what to strive for in their environment, they also learn to attend to some parts of that environment and not to others. One part of such adaptation is in learning search behavior, which we will consider in a moment. Here we wish to note two related, but different, adaptations:

  1. In evaluating performance by explicit measurable criteria, organizations learn to attend to some criteria and ignore others. For example, suppose an organization subunit has responsibility for a specific organizational goal. Since this goal is ordinarily stated in relatively nonoperational terms, the subunit must develop some observable indices of performance on the Among the indices objectively available to the subunit, which will be used? Observation suggests this is a typical case of learning. Subunits in the short run do not change indices significantly. However, there are long-run shifts toward indices that produce generally satisfactory results (i.e., in this case, usually show the subunit to be performing well).
  2. Organizations learn to pay attention to some parts of their comparative environment and to ignore other parts. We have assumed that one of the parameters in the goal adaptation function is a parameter reflecting the sensitivity of the organization to external comparisons. This parameter is not We would expect it to change over time as such comparisons do or do not produce results (in the form of goals) that are satisfactory to the important groups in the coalition. At the same time, we have represented by C in the goal adaptation function a summary description of comparable organizations. Concealed in such an abstract form is organizational learning with respect to what is properly comparable. With which attributes of which organizations should we compare ourselves? Although in a relatively short-run model we might reasonably consider this fixed, we would expect that in the long run we would require a model in which such attention factors changed.

Adaptation in search rules. If we assume that search is problemoriented, we must also assume that search rules change. Most simply, what we require in the models are considerations of the following type: when an organization discovers a solution to a problem  by searching in a particular way, it will be more likely to search in that way in future problems of the same type; when an organization fails to find a solution by searching in a particular way, it will be less likely to search in that way in future problems of the same type. Thus, the order in which various alternative solutions to a problem are considered will change as the organization experiences success or failure with alternatives.

In a similar fashion, the code (or language) for communicating information about alternatives and their consequences adapts to experience. Any decision-making system develops codes for communicating information about the environment. Such a code partitions all possible states of the world into a relatively small number of classes of states. Learning consists in changes in the partitioning. In general, we assume the gradual development of an efficient code in terms of the decision rules currently in use. Thus, if a decision rule is designed to choose between two alternatives, the information code will tend to reduce all possible states of the world to two classes. If the decision rules change, we assume a change in the information code, but only after a time lag reflecting the rate of learning. The short-run consequences of incompatibilities between the coding rules and the decision rules form some of the more interesting long-run dynamic features of an organizational decision-making model.

Source: Skyttner Lars (2006), General Systems Theory: Problems, Perspectives, Practice, Wspc, 2nd Edition.

Leave a Reply

Your email address will not be published. Required fields are marked *