Decision Making Under Risk in Organisations

The Case of German Waste Management

Eckard Kamper

• Publisher: Routledge
• Year: 2018
• eBook ISBN: 9781351730075
• Topic: Business
• Subtopic: Business General
• Index: Business

This title was first published in 2000. By comparing how two local authorities deal with the risks involved in Germany's environmental waste policy, this text questions how organizations deal with making decisions in situations of risk in general. Using a combination of risk sociology and institutional theories of organization, Kamper examines how organizations develop institutional structures to cope with risk-taking decision-making. In doing so, he challenges the commonly-held view that the most important factor required when making risky decisions is rationality, and instead argues that the key is an ability to absorb uncertainty. The volume seeks to show how organizations develop institutional structures to cope with risk-taking which both absorbs uncertainty and allocates responsibility, and how an understanding of social structures is crucial to understanding how such decisions are made by organizations.

1 Decision Making under Risk

The present study tries to analyse organisational processes of decision making under conditions of risk. The challenge for a sociological analysis of decision making under risk is, however, that the long tradition of theorising about problems of risk in decision making has taken place under the auspices of economic concepts of decision, risk, and uncertainty. Only in recent years, social theorists have put emphasis on developing genuinely sociological concepts of decision making, and of risk. These theoretical developments suggest that the classic conceptions of risk, rooted in economic theory, were not appropriate to guide empirical analyses of the social conflicts concerning risk (cf. e.g. Douglas 1985; Perrow 1984; Wynne 1987; Short 1984; 1989). Thus, for sociological analyses, it is necessary to develop concepts that deviate in almost all relevant aspects from the economic tradition.

In the following chapter, I will try to highlight the main characteristics of a sociological account of decision making under risk. This is necessary in order to clarify the deviations from the economic concepts. The economic tradition describes decision making as an action performed by individual or corporate actors who, on the basis of given preferences choose a course of action out of a given range of alternatives. In a sociological concept, decision making is an activity generated by behavioural expectations. The main consequence of this concept is that preferences lose their central theoretical position in explaining decision making processes. More important for the analysis of decision making are the expectations the decision reacts to.

The development toward this sociological approach will be sketched in the following sections (1.1, 1.2, 1.3). This general framework will then be applied to problems of decision making under risk (1.4).

1.1 Risk: The Initial Formulation

The initial formulation of a concept of risk was made by Frank Knight (1921). The context for the concept is provided by the framework of neoclassical economics. Within this framework, a decision situation can be characterised as follows: An actor is facing a situation with a range of decision alternatives. The actor is equipped with preferences, that are stable, exogenous, and transitive. Furthermore, the actor is characterised by maximising behaviour, i.e. he seeks to choose the alternative that serves his preferences best. The actor then makes his decision by evaluating the possible future consequences connected with the available alternatives. With the help of his preferences, the actor chooses the alternative that leads to the best consequences in terms of the actor's preferences. The interesting question for research on decision making processes is then the degree of rationality the decision maker achieves with his choice. Rationality is defined as the degree to which the decision maker manages to choose the alternative that serves his preferences best (Elster 1989; Luhmann 1988: 276).

Within this general framework, Knight places his concept of risk (1921: 197ff.). Risk is then to be distinguished from uncertainty. The actor finds himself in a situation of uncertainty if he cannot know whether the available alternatives really lead to the consequences he expects. This uncertainty is not measurable, because the actor does not know which factors influence the decision outcome. Risk, on the contrary, refers to measurable uncertainty. The actors thus know all of the possible decision outcomes, and the remaining problem is the calculation of the probability of their occurrence (cf. Bonß 1995: 98pp.). For Knight, only situations of risk can be subject to scientific analysis. The problem is how to turn uncertainties into risk which in turn is considered to be an information problem.

This basic concept of decision making under risk has been highly influential. It has mainly guided economic research (e.g. Arrow 1970; Anand 1993; Dickinson et al. 1994; Hey 1997). Since the sixties, when social conflicts about the risks of large technical systems arose, the concept has also been the basis for the theories of scientific risk assessment. In these theories, risk is defined as the product of the magnitude of a possible harm times the probability of its occurrence. The result is a quantifiable, one-dimensional, and probabilistic measure of risk that allows the comparison of different risks such as the use of nuclear energy, smoking, car driving, etc. This approach has proven to be extremely useful in improving the safety of complex technologies (cf. Bechmann 1993; Renn 1992).

However, from a sociological perspective, the approach to risk based on economic theorising has its deficits. First of all, one can easily see that situations of risk, as defined by Knight, will appear extremely seldom, since normally the decision makers will not be able to identify all the factors that influence the probability that certain decision outcomes will occur (Bonß 1995: 100). Second, the concept of a rational, maximising decision maker has been criticised, not only in the context of risk (e.g. Smelser 1992). For an organisational analysis, the concept requires that organisations be conceived as actors-a difficult demand (Hannan 1992; Jansen 1997; Kappelhoff 1997). Third, empirical research on risk taking behaviour has shown that probabilistic calculations are without any relevance for decision making under risk (March/Shapira 1988). Finally, the theories of risk assessment, based on Knight's concepts, have proved to be counterproductive in social conflicts about risky technologies. It has turned out to be impossible to fix a universal, objective measure of risk that is accepted independently from the social context (Renn 1992; Douglas 1985; Wynne 1987).

Accounts of risk within the economic tradition are therefore inappropriate for a sociological analysis of decision making under risk. In order to outline the sociological concepts that have been developed in the eighties and the early nineties, I will precede as follows in the coming sections. Firstly, I will concentrate on possible alternatives to the general concept of decision making (1.2). This will lead to the substitution of preference by behavioural expectation as the basic concept in a theory of choice (1.3). The starting point for this line of argument will be the behavioural theory of decision making, which contains many important elements of a sociological approach to decision making. In the last step, I will sketch the application of this general account of decision making to problems of decisions under risk (1.4).

1.2 Toward a Sociological Concept of Decision Making

The emergence of a specific sociological approach to problems of decision making can best be traced by examining the development of the behavioural theory of decision making, a branch of theory that is mainly connected with the names of Herbert Simon and James March (Simon 1957; March and Simon 1958; cf. Berger/Bernhard-Mehlich 1993). The key terms that characterise the development of the behavioural theory of decision making are bounded rationality (1.2.1), ambiguity (1.2.2), and decision making as rule following (1.2.3). This step-by-step rejection of the economic theory of decision making will be analysed in the following sections.

1.2.1 Bounded Rationality

The first important modification of the pure economic approach is the concept of bounded rationality. Exactly what this concept implies can be exemplified by the basic decision situation as sketched above. In contrast to a model of perfect rationality, the notion of bounded rationality introduces several limitations to a rational choice within this model. Firstly, the decision maker cannot know all possible alternatives that an external observer might identify for instance. Second, the decision maker has incomplete knowledge with respect to the consequences of each alternative. Third, it is very difficult to evaluate future events correctly, even if one has a consistent preference function.

These cognitive limitations in the process of decision making lead to the phenomenon that, in general, satisfactory instead of optimal decisions are made. This means that not all possible alternatives are examined, but only the first one that promises satisfactory results according to the underlying preferences.

For the central research question, the degree of rationality that is achieved by the decision, this means the following. Although an external observer might judge the behaviour of the decision maker as slightly irrational, the decision maker in his specific context at least intends to act rationally (March/Simon 1958; Smelser 1992: 388). If one takes into account the idiosyncratic conditions of the decision situation (complexity of the environment, structure of the organisation), one might even state that the decision maker achieves the highest degree of rationality possible under the given circumstances. Therefore, for the analysis of rationality in decision making behaviour, it is necessary to analyse the structural context in which the decision takes place.

March and Simon (1958) argue that there is no objective rationality. One can only speak of rationality in relation to a certain frame of reference. This is what is meant by the term bounded rationality. In contrast to the pure economic model the ability of the decision maker to know all of the available consequences and to anticipate all of the consequences connected with the alternatives is questioned. The modifications are thus modest, with the problem of rationality remaining in the centre of interest.

1.2.2 Ambiguity

In his later works, especially in collaboration with Johan Olsen, James March formulated a more radical concept of decision making in organisations, that rejects assumptions of rationality to a very large extent. The basic concept of these studies (Cohen et al. 1972; March/Olsen 1976; March 1988) is the notion of ambiguity. The main consequence of this concept is that preferences lose their central position in explaining choices. Since rationality in the economic theories of choice is defined in relation to the decision maker's preferences, the notion of ambiguity makes it difficult to speak of rational decision making.

Ambiguity in organisations refers to three factors. Firstly, the organisational technologies are unclear. Organisational technology in this context means knowledge of how to transform an input into an output. There are organisations with clear technologies, for example a producer of cars who transforms delivered parts into automobiles, and there are organisations with extremely unclear technologies, for example a university who has to transform young people into scientists. Organisations facing completely new tasks, for instance parts of public administration implementing a new political programme, also suffer from this lack of knowledge of what they are actually doing. The second aspect of ambiguity is that the goals and preferences are ill-defined, inconsistent and unstable. Third, there is no stable participation of decision makers in the decision process and the attention of the participants to the decision fluctuates.

These three factors generate conditions of ambiguity. The existence of unclear technologies and preferences and the fluctuating attention of the participants to the situation have serious impacts on the nature of the decision process and on the achievable level of rationality.

In order to demonstrate the consequences of conditions of ambiguity for organisational decision processes, March and Olsen assume a four-step cycle of choice: (1) Individual cognitions and preferences, (2) individual action and participation in organisational decisions, (3) organisational choices, (4) events in the environment of the organisation. (March/Olsen 1976: 13) A complete cycle of choice would imply that individual preferences are the basis for individual actions in the organisation, which then lead to organisational choices. The environmental consequences of the organisational choices are then observed by the decision makers and change or confirm their preferences, that are then the basis for further actions. This complete cycle of choice would suggest a certain degree of rationality as it allows the organisation to decide, to observe the consequences of the decision and then to react by improved actions. The result would be an adaptive rationality as suggested by Cyert and March (1963: 117): The organisation would not necessarily act rationally in the classical sense, but it would be able to learn from experience.

However, the above mentioned conditions of ambiguity inhibit the cycle of choice on each of its stages. First of all, due to the fluctuating attention and participation of the members in decision processes, the individual cognitions are not always the basis for the actions they contribute to the organisational processes. There are actions in organisations which are not based on individual preferences, and many preferences are not transformed into actions. Second, the coupling between individual actions and organisational choices is quite loose. Third, the connection between actions and environmental events is far from being clear. This is due to the fact that environments of organisations are complex and that the organisational technologies are often not well understood. 'As a result, the same organisational action will have different responses at different times; different organisational actions will have the same response.' (March/Olsen 1976: 17) Finally, the connection between events in the environment and the cognitions of the individual member is a complex one. Observed events have to be interpreted, causal relationships are not easy to identify.

After introducing the concept of ambiguity, the theory allows a more radical critique of concepts of rational decision making. The concept of bounded rationality questioned the possibility of knowing all alternatives and of anticipating the future consequences of the alternatives correctly. The notion of decision making under ambiguity goes even further: The preferences in organisational choice are inconsistent, contested, and unstable. Often, they cannot guide the decision making and learning behaviour in the organisation. The evaluation of future consequences in decision making, for which preferences are the basis, becomes highly ambiguous and even the correct interpretation of past events, the essential precondition of learning from experience, becomes very difficult. The modification of the concept of preferences enables the behavioural theory of organisational decision making to deconstruct expectations of rational choice in a very serious way. If it is correct that preferences are unclear, unstable and endogenous, i.e. that they change during the process of deciding and of evaluating the consequences (March 1988), some basic assumptions of theories of rational choice are challenged.

As the preferences of the decision makers lose their position as the central explaining factor for choice, it is difficult to define rationality in this context. Decisions often take place without clear underlying preferences, thus it is difficult to analyse whether the decision maker has chosen the alternative that serves his preferences best, or in other words, whether he has acted in a rational manner.

The question is, however, which concept could substitute that of preference in a non-economic theory of decision making. If one takes seriously the problems of ambiguity in organisations, one cannot keep 'preferences' as the basic variable to explain decision making behaviour. Conditions of ambiguity de-couple preferences and choices; thus preferences cannot explain the decisions that can be observed.

An answer is prepared by the development of the behavioural theory of decision making toward an institutional approach, which is mainly applied to political decisions. This perspective has been outlined by March and Olsen (1989; 1995) and suggests the concept of institutions as the central one for a theory of decisions in organisations. This approach will be sketched in the following section. Since the term 'institution' is however quite ambiguous, I will argue for modifications in the concept of institutions. These modifications, borrowed from the systems theory of Niklas Luhmann, allow for a framework for the analysis of decision making that does not need preferences as a basic concept. The new central term is that of behavioural expectations.

1.2.3 Decision Making as Rule Following

March and Olsen (1989; 1995; cf. March 1994) argue that there are two possible logics of action that underlie processes of decision making. They call them logic of consequences and logic of appropriateness. To argue, as March and Olsen do, that decisions take place according to a logic of appropriateness and not to a logic of consequences, implies a further rejection of expectations of rationality, as defined in economic theory.

Conceiving of actions and decisions as determined by a logic of consequences implies the following aspects: Decision alternatives are chosen by evaluating their perceived consequences according to the preferences of the decision maker. This logic of action is directed towards the future and it is characterised by anticipation, calculation, and analysis.

The logic of appropriateness is a rule-based one. Decisions take place by matching situations to existing rules of the organisation. As this logic of action refers to organisational rules that have been developed for past events, it is more directed to the past. It is not a logic of anticipation but one of recognising certain situations and classifying them as familiar or as new.

It is not only theories of decision making working with a concept of rational choice, which at least implicitly work with the assumption that decisions follow a logic of consequences, but also the behavioural theories of organised decision making, especially the contributions of March and Simon (1958) and Cyert and March (1963). Although these theories assume serious restrictions to anticipatory actions, actors still try to choose a course of action by anticipating its possible consequences. If one assumes that actions mainly are chosen by their appropriateness to the situation, the picture of decision making in organisations becomes different.

March and Olsen (1989: 21; 1995: 28) argue that political decisions in particular are made by referring to the logic of appropriateness. Political organisations are particularly prone to conflicting demands and expectations, and these contradictions can best be handled by the symbolic power of appropriate rules (March/Olsen 1989). It is this point in the theory, where the notion of an institutional perspective on decision making becomes important. What is appropriate to a given situation is predetermined by institutionalised structures, procedures, routines, identities, and role expectations. March and Olsen use the term institution as follows: 'We use the term in a more general sense to refer not only to legislatures, executives, and judiciaries but also to systems of law, social organization (such as the media, markets, or the family), and identities or roles (such as 'citizen', 'official', or 'individual').' (March/Olsen 1995: 27)

The notion that the decision situation is structured by institutions in the above sense has important consequences for a concept of decision making. For example uncertainty in decision making does not refer to uncertainty about future consequences or about the own preferences; ...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Contents
• Preface
• Acknowledgements
• 1 Decision Making under Risk
• 2 Risk and Organisations
• 3 Risks of Waste Management
• 4 Risk and Participation
• 5 Risk and Routines
• 6 Summary and Conclusions
• Bibliography