The allocation of resources to research and development in the firm

Abstract

Introduction The study and analysis of economic aspects of technological change is a fairly novel preoccupation of economists. Until recently there has been a conspicuous disregard of this topic by economic theorists, despite its recognised importance in industrial competition and economic growth. As Jewkes et al (1969) point out, future historians will no doubt find it remarkable that so little systematic analysis was conducted in this area in the first half of the twentieth century. While the post-war period has seen a gradual development of active interest on "the part of economists, it still remains a comparatively neglected topic. The most obvious and possibly most important reason for this neglect is the difficulty involved in adapting and applying conventional economic theory to this area. Innovation is, to a greater or lesser extent, a venture into the unknown as far as its development is concerned. In such circumstances past experience and quantitative techniques may provide minimal guidance for decision-makers. It is in this context, and with due acknowledgement of the difficulties faced in model-building in this area, that an approach to analysis of resource allocation to research and development in the corporation is developed later in the thesis. In particular, R & D budgeting techniques, possible determinants of the level of R & D expenditure, and the distribution of resources to basic research, will be studied in later chapters. It is hoped the hypotheses tested in this respect will contribute towards an improved understanding of the nature of technological change in the modem corporation. The thesis is divided into part I, consisting of chapters 1 to 5 inclusive, and part II, consisting of the last four chapters. Part I is concerned with the development of a model of corporate decision-making, the model itself being formulated in chapter 5. The early chapters lay the ground for this development by dealing with three related topics on corporate R & D. Chapter 2 deals with the characteristics of research and development at project level, particularly the relationship between science and technology, and the pervasiveness of uncertainty in R & D work. In this latter respect it provides a basis for criticism of the neoclassical and statistical theories discussed in chapter 3. However chapter 2 also develops the concept of hierarchical arrangement, or ordering, of R & D sub-systems which is of use later in analysing the determinants of basic research activity. Chapter 3 discusses the problems of theory application in conditions of pervasive uncertainty. Three of the approaches discussed - neoclassical economics, decision making under uncertainty, and behavioural theory, have a common bond in that they were initially developed and applied to problems other than research and development, but each has subsequently been suggested to be applicable to problems of R & D and technological change. A fourth economic approach developed by Penrose (1959) to deal with the growth of firms is also discussed, partly because of its potential application to the area of technological - change, but also because it provides useful guidelines for subsequent theory building, in conjunction with the behavioural theory of the firm. Chapters 2 and 3 together are intended to demonstrate the difficulties of applying received theory to technological change. Criticism is directed to theory application in this specific area and is not intended to be general criticism of the theories as such. It is in this context that chapter 4 takes a wider look at corporate decision-making and resource allocation, with special reference to the role of technological change in this framework. The typically hierarchical nature of corporate resource allocation is pointed out, and the role of R & D as a specialised and institutionalised function in the modem corporation is argued. It is suggested corporations must be regarded as open systems "which maintain themselves through constant commerce with their environment, i.e. a consistent inflow and outflow of energy through permeable boundaries," (Katz and Kahn, 1966, pp.18-19). In this thesis this is interpreted to mean that decision makers not only react to the corporate environment, but consciously and autonomously act to shape and mould the environment itself. This provides the basis for the model development in chapter 5. The arguments developed in the previous chapter contribute to the model of the firm as a hierarchically organised open system in which R & D operates as a specialised function. Part I, then, is concerned with the development of the model of the firm as a hierarchically organised open Introduction The study and analysis of economic aspects of technological change is a fairly novel preoccupation of economists. Until recently there has been a conspicuous disregard of this topic by economic theorists, despite its recognised importance in industrial competition and economic growth. As Jewkes et al (1969) point out, future historians will no doubt find it remarkable that so little systematic analysis was conducted in this area in the first half of the twentieth century. While the post-war period has seen a gradual development of active interest on "the part of economists, it still remains a comparatively neglected topic. The most obvious and possibly most important reason for this neglect is the difficulty involved in adapting and applying conventional economic theory to this area. Innovation is, to a greater or lesser extent, a venture into the unknown as far as its development is concerned. In such circumstances past experience and quantitative techniques may provide minimal guidance for decision-makers. It is in this context, and with due acknowledgement of the difficulties faced in model-building in this area, that an approach to analysis of resource allocation to research and development in the corporation is developed later in the thesis. In particular, R & D budgeting techniques, possible determinants of the level of R & D expenditure, and the distribution of resources to basic research, will be studied in later chapters. It is hoped the hypotheses tested in this respect will contribute towards an improved understanding of the nature of technological change in the modem corporation. The thesis is divided into part I, consisting of chapters 1 to 5 inclusive, and part II, consisting of the last four chapters. Part I is concerned with the development of a model of corporate decision-making, the model itself being formulated in chapter 5. The early chapters lay the ground for this development by dealing with three related topics on corporate R & D. Chapter 2 deals with the characteristics of research and development at project level, particularly the relationship between science and technology, and the pervasiveness of uncertainty in R & D work. In this latter respect it provides a basis for criticism of the neoclassical and statistical theories discussed in chapter 3. However chapter 2 also develops the concept of hierarchical arrangement, or ordering, of R & D sub-systems which is of use later in analysing the determinants of basic research activity. Chapter 3 discusses the problems of theory application in conditions of pervasive uncertainty. Three of the approaches discussed - neoclassical economics, decision making under uncertainty, and behavioural theory, have a common bond in that they were initially developed and applied to problems other than research and development, but each has subsequently been suggested to be applicable to problems of R & D and technological change. A fourth economic approach developed by Penrose (1959) to deal with the growth of firms is also discussed, partly because of its potential application to the area of technological - change, but also because it provides useful guidelines for subsequent theory building, in conjunction with the behavioural theory of the firm. Chapters 2 and 3 together are intended to demonstrate the difficulties of applying received theory to technological change. Criticism is directed to theory application in this specific area and is not intended to be general criticism of the theories as such. It is in this context that chapter 4 takes a wider look at corporate decision-making and resource allocation, with special reference to the role of technological change in this framework. The typically hierarchical nature of corporate resource allocation is pointed out, and the role of R & D as a specialised and institutionalised function in the modem corporation is argued. It is suggested corporations must be regarded as open systems "which maintain themselves through constant commerce with their environment, i.e. a consistent inflow and outflow of energy through permeable boundaries," (Katz and Kahn, 1966, pp.18-19). In this thesis this is interpreted to mean that decision makers not only react to the corporate environment, but consciously and autonomously act to shape and mould the environment itself. This provides the basis for the model development in chapter 5. The arguments developed in the previous chapter contribute to the model of the firm as a hierarchically organised open system in which R & D operates as a specialised function. Part I, then, is concerned with the development of the model of the firm as a hierarchically organised open system in which R & D operates as a specialised function. Part II applies this open system interpretation in the empirical analyses of chapters 6 to 8 inclusive. Chapter 5 is concerned with accounting for dissimilarities in budgeting conventions adopted by corporations operating under different circumstances in Western Europe and the United States. The evidence of a number of surveys and studies is considered in this chapter, and it is suggested that not only does the open system interpretation reconcile apparently arbitrary differences in budgeting "style", but also that the systems interpretation developed here provides a rational basis for rule-of-thumb budgeting techniques employed by many large corporations, and frequently described as "illogical" or "irrational". In chapter 7, hypotheses based on the open systems framework are developed, and regression analysis conducted in an empirical examination of the hypotheses. Specifically, possible determinants of R & D and basic research activity in U.S.industry are investigated. As well as constituting an empirical study of the possible influences on corporate allocations to technological change, it is suggested that the model developed in part I provides a sound framework for the empirical hypotheses of this chapter. Not only does it avoid many of the conceptual difficulties of conventional approaches such as neoclassical theory, it also illustrates how the hierarchic "top-down" system of resource allocation widely adopted by large corporations may be interpreted as rational behaviour. Chapter 8 is concerned with intra-industry variation in allocations to technological change activity, and may be regarded as complementing the essentially industry level orientation of chapter 7. It discusses the role of rivalry as far as competition in innovative activity is concerned, in particular the propensity of corporations to match or imitate competitor allocations, e.g. in terms of percentage of sales allocated to research and development. Apparently contradictory evidence as to whether or not competitive matching is a prevalent form of industrial behaviour is analysed in this chapter, and a reconciliation is suggested based on the adaptive learning aspect of the systems approach developed earlier. Part II is concluded with a short summary as to the main conclusions of the thesis and possible implications for future analysis. It is suggested that neoclassical economics does not provide an adequate framework for investigation of technological change, and that the alternative approach developed here and based on concepts developed in general system theory may generate a more satisfactory basis for study of certain aspects of this problem area. While the difficulties of applying neoclassical economics in this area are generally agreed, it is hoped the potential usefulness of general system theory is demonstrated through the studies discussed in chapters 6 to 8 inclusive. It will be emphasised, however, that the usefulness of the systems approach can only be examined indirectly. There are no precise, testable hypotheses provided. As Katz and Kahn (1966) comment; "In some respects open-system theory is not a theory at all; it does not pretend to the specific sequences of cause and effect, the specific hypotheses and tests of hypotheses which are the basic elements of theory. Open-system theory is rather a framework, a meta-theory" (p.452). The approach provides a frame of reference within which lower level hypotheses capable of empirical testing can be generated. The concepts and interpretations of the systems approach provides a perspective and basis for empirical analysis, not a set of ready made hypotheses. Thus, rejection of a lower level hypothesis need not imply rejection of the systems approach. This apparent irrefutability of the systems approach in no way invalidates its use. Most theoretical approaches incorporate variables and relationships at higher levels that are not directly observable, but which have operational correlates at lower levels; the relevance or otherwise of the systems approach will be debated on the basis of the performance of lower level hypothesis. As far as the empirical hypotheses consistent with alternative theoretical approaches are concerned, hypotheses may be similar or conflicting between these different frameworks, or it may be that a hypothesis may have no corresponding or conflicting hypotheses in other approaches. The latter is particularly likely if one approach suggests a rich vein of testable lower level hypotheses. These possibilities are of particular relevance as far as the regression analysis conducted in chapter 7 is concerned. Some of the hypotheses may be consistent with the neoclassical theory of the firm, while with others it is difficult to see how they might be developed in a project based theory of the firm framework. It will be argued that the systems approach must be judged on assessment of the lower level hypotheses considered as a whole. The regression analysis of chapter 7 appears to offer a good explanation, not only of the behaviour of R & D at industry levels, but also characteristics of uptake of basic research activity by corporations. Although conventionally it is generally assumed that resource allocation to this latter phenomena is particularly difficult, if not impossible, to account for in economic analysis, chapter 7 suggests that this may not be the case. Not all the lower level hypotheses are supported by the evidence in this particular study but it will be argued that the general performance of the regression analysis is good. The thesis is therefore intended to provide a useful though partial framework for the analysis of aspects of technological change in the corporation. One area of obvious relevance which is virtually neglected is that of selection of projects and resource allocation within the R & D budget constraint; however it is argued later that this is justified in terms of the points made in chapters 4 and 5. Bearing in mind such potential restrictions on the analysis, a main aim of the thesis is the provision of a useful conceptual framework for the analysis of resource allocation to technological change in the large modern corporation, and demonstration of the potential applicability of rational analysis to areas where it has been frequently suggested to be irrelevant. However there is a second main objective of this thesis which evolved from arid was stimulated by, the nature of the problems encountered in developing this framework. It will be argued that the emphasis on individual elements in the resource allocating process in analytic theories such as those of the neoclassical economists may actually observe or inhibit understanding of decision making in the corporation. Neoclassical theory has as its building blocks the individual consumer, product, and projects, larger units being defined in terms of sums of individual elements. The technique of aggregation is used in moving from micro-levels to higher levels of analysis. It is a basic tenet of this thesis that reduction to component elements and definition of higher levels as aggregates may be inappropriate in certain circumstances and for investigation of specific phenomena. Instead a systemic or holistic view of the corporation is suggested as an alternative approach to aspects of resource allocation. Two main reasons are given for this. Firstly, advantage may be taken of redundancy at lower levels; only limited and highly abstract detail may be necessary to adequately describe or approximate a systems behaviour. This is illustrated in a pragmatic way by simulation models which in certain circumstances may provide a good description of the behaviour of complex systems while utilising only highly selective and schematic information. Secondly, system description and behaviour may be non-reducible as far as specification in terms of constituent elements is concerned. Pattern, configuration or "gestalt" may be established at relatively high levels of abstraction, and may not be directly derivable from consideration of components alone. Both these arguments are used in interpreting the corporation as an adaptive, hierarchically structured system, and will be developed at greater length in subsequent discussion. Thus the objectives of the thesis are basically twofold. Firstly, it is intended that a satisfactory and useful approach to sane decision-making problems relating to technological change in the modern corporation may be developed. Secondly, it is hoped that a convincing case may be made for the argument that the dominance of a reductionist perspective in standard economic approaches may hinder rather than assist model building in some cases. It is hoped the following analysis may demonstrate the possible usefulness of an alternative economic model of the firm based on the constructs of general system theory.