It reflects a modern concern to establish a more systematic understanding of the process of creative destruction

The central concern of this paper is economic growth under the rules of restless capitalism, rules under which multiple, uncoordinated innovative activities are ordered by market processes to produce patterns of growth and development in the economy as well as in its framing institutions. The central theme of this worldview is that growth follows from sequences of technical, organisational and institutional changes that create and absorb new areas of productive activity and consumption into the economic structure. From this angle, growth is the result of the exercise of imagination within the frame of a market system. These developments arise from within the system, and they are grounded, ultimately, in changes in private knowledge and public understanding (Metcalfe, 2001b; Potts, 2001). Thus, capitalism is restless because knowledge is restless; there never can be any equilibrium in respect of knowledge and, consequently, the development of an economy is unpredictable and open-ended. The social and institutional dimension of knowledge is as important as the economic dimension to any understanding of this dynamic and the decline of activity is as important as is expansion of activity in mapping the contours of development. This is a theme, which is central to the Nelson and Winter project as first fully articulated in their An Evolutionary Theory of Economic Change. It reflects a modern concern to establish a more systematic understanding of the process of creative destruction (Schumpeter, 1944). However, it is a perspective that contrasts sharply with modern growth theory, endogenous or otherwise in which structural dynamics are eliminated either by assuming a single commodity or by the assumption of semi-stationary or proportional expansion in which all sectors expand at the same rate (Pasinetti, 1993). Such a macro approach has averaged away not only the facts of the uneven incidence of growth and technical change across sectors, it has also hidden from view the very processes by which changes in technique, organisation and institution are originated and come gradually to have their economic and social effects. Furthermore, this approach inevitably losses sight of how these processes of transformation shape the subsequent growth of practical knowledge and give to the growth of knowledge in general a powerful autocatalytic dynamic in which the solution of problems defines yet further problems. In the Nelson and Winter picture, by contrast, growth is very much bottom up but to interpret their argument entirely in this light is as misleading as to see it only in a top down fashion. Rather, the aggregate structures and patterns that emerge at higher levels in their evolutionary framework create important feedback effects to shape the underlying processes of microevolution of technique, organisation and institution. This inherent complexity of the interaction between technical progress and economic growth in modern economies means that simple categorisations of the growth process will be difficult to find. My conjecture is that Nelson and Winter have it right, the best way to make progress is to see this interaction as an adaptive evolutionary process, a triple, connected process of variation, selection and development operating at multiple levels. Evolutionary theory is naturally growth theory but it is the diversity in rates of growth, within appropriately defined populations, that is the focus of attention. The question is “Why growth rates differ and vary over time?” not “Why are they uniform and invariant?”. Structural change follows from diversity in growth, and the mutual determination of those growth rates means that they are emergent phenomena arising from replication and interaction. As Pasinetti (1981, 1993) rightly concluded, it is the presence of endogenous structural change, which makes it impossible to conduct the analysis of growth by macro economic methods alone [2][2]Pasinetti (1981), p. 65. It is a point with which Schumpeter…. Thus, our central theme, economies are transformed over time by the generation and application of new knowledge and we more or less imperfectly capture these qualitative and quantitative changes in the conventional measures of economic growth: transformation comes first and growth is the derivative consequence. [3][3]The idea of growth as the consequence of transformation is…
2The fact that knowledge based systems cannot be in equilibrium if knowledge is not in equilibrium, is the basis for a “far from equilibrium” judgement about the economic process. Every position generates its own destruction, which is exactly the point of Schumpeter’s emphasis on development from within the system. Yet capitalist economies show remarkable evidence of order and pattern in the allocation and definition of resources and activities, which we naturally trace back to the co-ordinating role of markets. An economy without equilibrium is easy to conceive, a capitalist economy without co-ordination is not. Yet, the role of markets is not simply to be equated with the properties they might have in relation to the efficient allocation of given resources over time. Markets are important as loci of knowledge generation, and in being open to and indeed generating the incentives for endogenous change. It is the openness of markets to the challenge of established positions that gives innovation and the entrepreneur such a powerful role to play in economic transformation. This is the Schumpeterian connection to this paper; it is equally a connection with broader Austrian and Hayekian themes about economic transformation as an open-ended discovery process. [4][4]The exploration of this point would take us beyond the confines… The evolutionary view underpinning this paper locates growth within market processes and the framework of institutions that defines modern capitalism. It emphasises the open nature of competition within these frames, and it gives innovation, in the broad, pride of place as the primary stimulus to and consequence of the growth and development of economies.
3There are many varieties of evolutionary growth theory but the variant that I develop here is based on the theory of niche formation and exploitation as developed by evolutionary ecologists and I shall use these ideas to explore some economic precursors of an evolutionary perspective on endogenous growth. For it turns out that, prior to the Keynesian revolution and Harrod’s formulation of modern aggregate growth theory in the late 1930s, there was in place a rich empirical and theoretical literature on the problem of secular economic change, a literature that posed this problem in terms of the nature of innovation and the development of technology. This literature has at least four main branches of which only the first three are firmly within economics. First, is the line of analysis established in Schumpeter’s Theory of Economic Development, with its emphasis on the entrepreneur, new business formation and introduction of new forms of economic activity and their subsequent spread within an economy. An ecologist would readily recognise this as a model of speciation and competition for economic niches. The important aspect of Schumpeter’s analysis is that it brings together stability in the capitalist order with instability in the capitalist system. The continuous transformations in economic form are associated with the creation and application of new combinations that arise from within the otherwise relatively more stable order of overarching institutions (Schumpeter, 1928). As he expressed the point,
4
“… what we unscientifically call economic progress means essentially putting productive resources to uses hitherto untried in practice, and withdrawing them from the uses they have served so far” (op. cit., p. 378, emphasis in original).
5Anyone familiar with modern accounts of evolving systems will recognise in this quotation the importance of linking innovations to the introduction of non-average behaviours and the opening up of new spaces for economic activity (Allen, 2001). Since mechanical systems can only be analysed in terms of average behaviours and processes, one can see immediately the significance of Schumpeter’s method as an early, if perhaps unwitting, exposition of evolutionary dynamics and a celebration of the significance of unrepresentative behaviour. Knight (1935) expressed similar views, in his critical account of the state of growth economics circa 1930 and the inappropriate attempt to place growth within a framework of equilibrium thought.
6No doubt, the basis of Schumpeter’s model of capitalist development is to be found in his knowledge of industrial history. Business Cycles (1939), as one example, contains a detailed account of how the first three decades of the 20^th century were a period of enormous economic transformation both qualitative and quantitative. This period witnessed the creation of major new dimensions to economic space in terms of such activities as the production of automobiles and the complementary industries, the spread of electricity in home and workshop and the production of new synthetic materials (Sayers, 1950). For Schumpeter, the idea of the steady growth of capitalism must surely have been a contradiction in terms. [5][5]It might perhaps be argued here that Schumpeter’s circular flow…
7While Schumpeter provided new perspectives to comprehend the economic processes at work, it was the detailed empirical investigations of Simon Kuznets and Arthur Burns, on secular trends in real output and the so-called retardation hypothesis at the level of individual industries, that provided a more sharply focused empirical foundation to an innovation based account of growth. Their ideas where explored further by Fabricant (1942) and, in the post war period, by Stigler (1947) and Gaston (1961). However, by then the aggregate orientation of growth theory was taking a firm hold and this second strand in our schema was rapidly passing into history. The third strand we find in a quite different and older tradition, the argument of Allyn Young (1928) to the effect that the link between technical progress and economic change was a deeper reflection of the Smithian argument on the extension of the division of labour. That is to say, the rate of advance of knowledge and thus economic progress is connected closely as cause and effect with the rate of growth of the market. Consequently, technical progress induces further advances in technique. All of these contributions lie firmly within the economics discipline and they singly and together form the basis for an evolutionary account of economic growth. Our fourth contribution is, by contrast, from a different domain and it concerns the background debate on “laws of growth” arising from within the emerging discipline of mathematical ecology in the 1920s but which had much wider implications for the study of industry and the economy. Lotka’s treatise Elements of Physical Biology, published in 1924 provided the definitive statement of this literature, and there he defines evolution in terms of the redistribution of the components of a system, that is, as structural change. This literature developed naturally into the analysis of a growth curves in different populations from which the extension to economic phenomena followed with ease (Prescott, 1922; Bratt, 1936; Vianelli, 1936; Windsor, 1932). [6][6]Ultimately, it emerged again, in different form in the idea of…
8Taking all four strands together connects the analysis of economic growth with a central topic in ecology, namely, the way in which a new species grows into its environmental niche. The insight here is that the dynamics of growth are governed by two considerations, the intrinsic rate of increase of the species (r) and the carrying capacity of the species in the environment (K) (Roughgarden, 1979; Slobodkin, 1980; May, 1974; Kingsland, 1985; Renshaw, 1991). This so-called theory is undoubtedly crude and oversimplified but it can be developed and applied in a variety of ways which lay bare the main features of the process of innovation based growth and transformation. [7][7]Andersen (1994) and Saviotti (1996) are two very cogent…
9My purpose in this essay is to retrace the argument back to its origins in the 1930s in the work of the “retardationists”, Kuznets and Burns in particular. In the process, we shall develop an economic interpretation of carrying capacity and intrinsic rate of increase so that their characterisation is not arbitrary but is related to the dynamics of the market process. It is clear that the literature I have referred to provides an endogenous account of economic growth. New knowledge defines new combinations, entrepreneurship introduces these new combinations into the space of economic activities, and those that pass the test of economic and social viability may spread further into the system attracting resources and demand and so enhancing or destroying the markets for existing activities. In the process, new knowledge is gathered, new opportunities emerge, improvements are made and so the process feeds on itself in autocatalytic fashion. However, there is no metric through which knowledge can be reduced to a scalar quantity. Knowledge is an ensemble not an aggregate, a union of individual’s knowledge not a sum, and it is quite inappropriate to think of it as some homogeneous substance that has an independent macro economic existence. To think in these terms is to suppress the very processes that create growth through the opening up and exploitation of new economic spaces. This is the evolutionary import of the work of Kuznets and Burns, and of Schumpeter and Young, to which we now turn.