Decisions, Technology and Organization

Audley Genus

• Publisher: Taylor & Francis
• Year: 2017
• Print ISBN: 9781138730618
• eBook ISBN: 9781351745819
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

This title was first published in 2000. This text analyzes the problems of managing large-scale technology projects. It addresses the contrast between projects' hoped-for benefits and the optimism with which they are promoted, and the high human, economic or environmental costs of their operation. The book is multi-disciplinary in its approach and an integration of different levels of analysis. It contains case studies that are analyzed in terms of the literatures on decision theory, public administration and strategic and project managment.

PART I

INTRODUCTION

1 Perspectives and Prescriptions: the Analysis of Large-scale Technology

Introduction

This book is about the theory and practice of decision-making and the construction of technology. It speaks to the poverty of attention to failed technological innovation identified by Pinch and Bijker (1987), which persists. It contributes to the literature on ‘grand-scale’ projects. Such projects provide society with some of its greatest enduring technical achievements but also with some of its most damaging or costly mistakes. Current illustrations of the type of ventures giving cause for concern include defence projects, which often incur very large cost overruns and long development delays (National Audit Office, 2000). Other contemporary projects such as dam-building schemes in Kurdish Turkey, the Oresundsbron road bridge linking Sweden and Denmark, or the Millennium Dome and Millennium Bridge further illustrate the potential ambivalence of large construction projects in terms of serving society. Critics attempt to sound their notes of caution. Supporters and promoters, however, point to the positive implications of these expenditures in terms of their contribution to domestic and foreign policy objectives. In the public debate, either heightened controversy ensues leading to inertia, or in the absence of criticism risky technology is ‘pushed’. These observations are hardly new. Yet, surprisingly, there is relatively little research on public policy towards and the operational management of this type of venture. The study of grand projects is an area of research in need of further scrutiny, therefore (Shapira and Berndt, 1997).

In particular, it is necessary to examine what is meant by failure and success and to account for the factors promoting successful or unsuccessful large-scale technology projects. Some clues on the matter may be obtained by considering the merits of rival prescriptions for decision or policy-making, in conjunction with an analysis of technical, economic and organizational aspects of the major projects featured in the book. With this in mind, a further aim of the book is to bridge the gap between the prescriptive literature on decision-making, which tends to neglect the content and context of the phenomena in question, and histories and case studies of technology development, which eschew generalization in favour of description. To the extent that this occurs it will be possible to enhance our understanding of decision-making in complex or uncertain situations, the management and evaluation of grand projects, and the social construction of technology.

There are a number of concepts that are core to understanding processes of design, implementation and performance evaluation in these areas. As suggested in the previous paragraph these notions have their origins in disparate academic disciplines. In the following sections it is intended to provide an overview of some of these key concepts and contributions from the relevant literatures. Such a sketch will help to place the case studies reported and discussed in the remaining chapters in some theoretical context. Admittedly, the selection of material covered in the following sections may seem arbitrary or incomplete to some readers. However, the following groups of notions and subject areas seem most apposite to the concerns of this book: decisions and flexibility; incrementalism; and technology policy and project organization.

Decisions and Flexibility

The first concern of the book is to improve our understanding of decision-making in uncertain situations and to contribute to the enhancement of decision-making in practice. Having said this we need to be mindful of some issues pertaining to the study of decisions and decision-making. Firstly, it is quite possible that ‘decision’ gets in the way of understanding decision-making (Mintzberg and Waters, 1990). The point is made that researchers tend to look for episodes of decision-making, often at the ‘top’ of an organization and in so doing miss the subtlety of decision-making as an ongoing process typically shaped by a variety of interests at various levels of the organization (or outside it). In addition whilst Pettigrew (1990) considers research on organizational decision-making to be ‘theoretically and empirically additive’ a problematic issue concerns the assumption that ‘decision’ equates to intentional commitment with action to follow. On this view implementation is the end stage of a linear process, which follows problem definition, search, analysis and choice. Research on decision processes in practice observes its ‘recursiveness’ (Sharfman and Dean, 1997), its iterative nature (Quinn, 1980) and the significance of context and process (Pettigrew, 1991). Treating decision-making as a recursive and iterative process draws attention to the cycling back and forth that decision-makers make between stages of a non-linear process. Studies of incrementalism have demonstrated the non-linear nature of decision-making in practice, and are reviewed in the next section of this chapter.

One way to approach the subtleties of the relationship between content, process and the context of decision-making is to try to analyze ‘nondecision-making’ as well as to commitments and action which represent the traces of decision-making that are more likely to be recorded, or recordable. As connected with the work of Bachrach and Baratz (1962) this concept draws attention to the process by which the mobilization of bias results in some options being kept off the decision-making agenda and by which the influence of certain actors is undermined. This is a notion to which the following chapters will return, as the discussion of the development of large-scale technology projects proceeds.

The key contributions on these issues over the past fifty years have offered theoretical analysis and critique of the rationalist conception of decision-making. It has become the conventional wisdom to acknowledge the limits of rationality, our own bounded rationality and the politics of decision-making. However, from time to time we see major decisions, most notably in the area of technology policy that seem to rely upon the thinking of the rational planner. On one view of this world, objectives are clear and constant, utilities transparently calculated, the problem confronting decision-makers (or more usually the individual ‘decision-maker’) is well-specified and incontrovertible. Data analysis can be ‘thrown at’ this problem to identify and to evaluate options leading to the selection of the optimum solution. Such an approach to decision-making may rely upon confident forecasts (of demand, costs and revenues) upon which sophisticated infrastructure will be installed. On another view the comprehensiveness of the analysis is relaxed. Instead of applying data to uncover the range of options open to the decision-maker, analysis is collected in support of one preferred option, to justify its selection (Carley, 1980). Whatever version of the model is employed this type of ‘synoptic’ rationality rests upon the assumption that

provided the right formula has been followed, the chooser can have confidence that whatever might happen in the future, there will never be a need to recognize [that] that original choice was erroneous and in need of revision, hence [there is] no point in searching for mistakes (Collingridge, 1992: 5).

The making of decisions in this manner contravenes what is known about the ordinary way in which human beings incorporate learning through low-cost trial and error into future decisions and actions. Fundamental to this concept of the decision-making process is the view that we all make mistakes and that these not only occur but, moreover, are unavoidable. The efficient detection and remedy of errors to allow a new course of action to be pursued may be frustrated by the ‘irreversibility’ (Douglas and Wildavsky, 1982, with regard to the control of risky technology) and also the ‘indivisibility’ of decisions (Schulman, 1975). To the extent that large-scale technology ventures can be said to have enjoyed success whilst breaching the rules concerning the adoption of approaches enabling efficient trial and error learning, a degree of comfort may be offered to policy-makers and managers operating in this manner. In contrast, the failure of such grand technology projects allows approaches valuing low-cost trial and error experimentation to be retained. Such approaches refer to need for flexible decision-making, discussion of which follows.

Flexibility

In general notions of flexibility refer to the capacity of decision-makers to modify strategies, policies or operations in the light of new circumstances or information, and at low cost. Genus (1995) considers flexible decision-making in terms of the degree of freedom of manoeuvre decision-makers have. The distinction has been made between the content of flexible decisions (referring to the extent to which alternative decision options are kept open) and the flexibility of decision processes (Sharfman and Dean, 1997: 192). Flexible decision processes are described as those that open up the decision process to new possibilities and ideas and away from stereotyped responses and routines. The distinction has also been made between information processing and ideological perspectives of flexibility (Sharfman and Dean 1997). In essence, research on flexible adaptation, which may be identified with the information processing perspective, focuses on the constraints to flexibility connected to decision-makers’ mental models and cognitive structures. The impact of these cognitive barriers biases the thinking of decision-makers and hence the range of actions they take. On the other hand, ideological factors linked to corporate culture or societal values shape decision-making in a number of areas including the awareness and definition of problems. In addition, the definition and legitimacy of objectives as well as the merits of alternative options may be skewed by core and shared beliefs and other normative influences. The distinction between information processing and ideological perspectives of flexibility may be considered in the context of the tension between competing explanations of the escalation of commitment to failing decisions (c.f. Drummond, 1999). Whereas a social-psychological explanation would focus on biased interpretations and ignorance of available information, along the lines suggested above (Staw and Ross, 1987), decision dilemma theorists look for factors that prevent better information being obtained about the risks involved (Bowen, 1987). These explanations of continued or escalating commitment to a course of action represent another line of discussion in the chapters below, regarding the sense of mission driving high profile projects and the barriers to managing them rationally. A further consideration concerns the constraints on future action associated with the specialized nature of assets required for large-scale technology investments. Here, Harrigan’s (1985) view of what she calls strategic flexibility refers to the barriers to exit from declining industries presented by a firm’s current investments. Such assets may be difficult to dispose of in a market short of potential buyers and so cause decision-makers to be ‘locked into’ previously taken decisions. This aspect of industry structure, she argues, may be exacerbated when the assets in question are dedicated to a specific use, or cannot easily be redeployed.

Incrementalism

The concept of incrementalism is most closely associated with the work of Charles Lindblom (1959; 1968; 1977; 1979; 1990; with Woodhouse, 1993). Lindblom makes the point that, understood in terms of the content and scale of decision-making, organizations rarely make ‘root’ changes to the status quo. The demands on information, our inability to compute from perfect information the optimum decision even if with the necessary data, and our preference to turn our backs on ‘ills’ suggests another model of decision-making which Lindblom calls the alternative the ‘branch’ or ‘incremental’ method. Here, only a limited number of courses of action marginally different from the status quo are considered as potential options for change. An important aspect of Lindblom’s view of incrementalism is that of seriality. In other words, decision-makers attack problems in an on-going way, such that incrementalism is a dynamic concept, and not a justification for conservatism or inertia. Further, it is an idea that prizes trial-and-error forms of learning, since the necessary experimentation that will bring forth innovation is done in a relatively small-scale way. In addition to the issue of the scale of decision-making, a second facet of incrementalism may be identified which relates to the structure and processes for decision-making. This embraces questions of the role of participation in and analysis for decision-making. Key is the emphasis on the politics of decision-making. Lindblom and Woodhouse (1993) argue that though imperfect, ‘intelligent democracy’ can attend to many of the difficulties of complex social problems with which analysis cannot cope (for a summary of limits to analysis see Table 1.1 below). Those interested in better social problem-solving should therefore focus on trying to understand what factors contribute to the effectiveness of intelligent democracy and which may facilitate its improvement.

Table 1.1 Limits of Analysis

a) poor quality/usability of policy research and data

b) fallibility:

- decision-making without political interaction can only take place where information and reasoning are sufficient bases on which to persuade parties

- information available is usually either of overwhelming or insufficient quantity or biased

- actors do not believe in the infallibility of information or experts

c) analysis cannot resolve conflicts of values

- e.g. what is the public interest?

d) time and cost

- in complex or urgent situations decisions cannot wait until ‘all the facts are in’; decision-makers draw on whatever is available

e) problem ‘formulation’

- problems to which analysis may be directed are not ‘given’; their priority ordering is not set either - these entail moral questions and issues of values and judgement. ‘The settlement requires politics more than analysis’.

Source: Lindblom and Woodhouse, 1993, chapter 2, pp. 20–22.

The factors that contribute to effective, intelligent democracy should not be confused with what seems to be the accepted wisdom concerning the success of democratic systems over totalitarian ones. This centres upon the responsiveness of policy-makers to popular control, either via the electoral system or freedom of speech, the press and so on. However, free societies are not necessarily capable of organizing to solve problems well and some democracies choose regimes that overthrow democracy. Instead, it has been observed that good social problem solving might be seen to rest upon certain key processes that are currently not used enough, with the effect that fundamental social problems persist over decades. Where policy does work it is when these processes are employed to good effect albeit in an unwitting manner. Overall, these ‘strategies’ allow such problems to become ‘far more manageable than could be achieved via analysis alone’ (Lindblom and Woodhouse, 1993: 24).

Briefly, the processes alluded to above include: partisanship; search for agreement; various forms of ‘strategic’ analysis; and the fragmentation of analysis among partisans. One of the defining features of the incremental approach is the recognition of the self-interest of partisans who pursue their own policy goals and have their own definitions of an issue. This recognition helps to side-step the old chestnut of how to make rational choices based on the best, unbiased analysis. Rather than to rely on analysis to facilitate an objective definition of a policy issue or its solution, partisanship may be thought of as ‘substituting politics for analysis’ (Lindblom and Woodhouse, 1993: 26).

The role of analysis in complex problem-solving is to facilitate the political interactions of partisan interest groups. The intelligence of the policy process is raised to the extent that partisans use data to persuade other protagonists: on the basis of reason; in ways which seek out what Quinn (1980) termed ‘zones of indifference’; or by appeal to the likely preferences of other parties. These processes may be considered more ‘intelligent’ processes of winning over rival viewpoints than the use of nonrational or irrational means represented by propaganda campaigns, vetoes, or bribes which do not require the persuasion of others of the merits of the option in question per se. The agreement or understanding that occurs as a result of these political interactions represents accommodations that a ‘working majority’ of partisans make regarding future actions. It is not necessary for there to be agreement on the definition of a problem or the facts of the matter. Indeed there may continue to be substantial differences in the formulations that various groups have of an issue and the options they favour for its resolution.

The essence of the incremental approach to the content of decision-making is that actions be modest rather than radical. The prescribed method for the analysis of decisions is for decision-makers to confine their deliberations to options varying only marginally from the status quo and the remedy of ‘egregious ills’ (Morone and Woodhouse, 1986). The view that decision-makers do and ought to engage with policy issues in a ‘serial’ rather than a comprehensive manner attracts criticism for tinkering with fundamental problems (most notably from Arrow, 1964; Dror, 1964; and Etzioni, 1967). The incrementalist’s retort to this suggestion is to challenge the critic to find instances where a ‘once-and-for-all’ approach has been demonstrably more effective than...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• List of Tables
• Preface
• Acknowledgements
• List of Abbreviations
• PART I: INTRODUCTION
• PART II: CASE STUDIES
• PART III: RETROSPECT AND PROSPECT
• Bibliography
• Index