Reviving and assessing the Kuznets law on innovation

Abstract

In the earlier years of the twentieth century economists were beginning to gain a deeper understanding of the nature of the economic change. Joseph Alois Schumpeter, for example, argued that the first key step in understanding economic change is to think carefully about innovation. When Kuznets began his work in economics in the mid 1920s economics had begun to develop the quantitative and mathematical dimensions that now characterise the discipline. In particular, Kuznets discovered what he argued was a second key step in understanding economic change, namely: the analysis of sectors where innovation actually takes place. He was a quintessential empiricist and adept at finding empirical regularities. It is the contention of this paper that Kuznets' early empirical work on sectoral growth enables us to formulate an empirical law on innovation. The paper is organised around a number of questions: What is the Kuznets law on innovation? Has this law been ignored? What is the relationship between the Kuznets law on innovation and the law of diminishing returns to innovative effort? What are the intrinsic limitations regarding the applicability of the Kuznets' law? The last part of the paper provides a summary and notes some limitations of the law.

Keywords:

• Innovation
• Economic laws
• Retardation of growth
• Diminishing returns to R&D effort

Notes

¹This paper is a preliminary summary of results published later in Kuznets (1930).

²‘Unbalanced’ because Kuznets stressed the primacy of facts over theory.

³This book (henceforth Secular Movements) is the result of the investigation undertaken by Kuznets while he was Research Fellow for the Social Science Research Council during the period 1925–1926. We note, in passing that the first chapter of Secular Movements, entitled “Retardation of Industrial Growth”, is virtually a replica of Kuznets (1929) with some additional material. For this reason all quotations are extracted from Secular Movements.

⁴This does not mean to imply that Kuznets' argument is faulty or inconsistent. It simply means that the major trends in his argument are difficult to disentangle. The reader is asked to bear in mind that Kuznets was in his late twenties and economics was surging on in apparent disorder at that time.

⁵The standard formulation of the law of supply and demand is as follows: ceteris paribus, the price tends to fall with excess supply and increase with excess demand.

⁶The primacy of pure theory without facts was probably induced by Friedman's methodological position, namely: the realism of the basic assumptions does not matter, what matters is the empirical validity of the conclusions.

⁷True advance can be achieved only through an iterative process in which theoretical formulation raises new empirical questions and the answers to these questions, in their turn, lead to new theoretical insights. The “givens” of today become the “unknowns” that will have to be explained tomorrow. This, incidentally, makes untenable the admittedly convenient methodological position according to which a theorist does not need to verify directly the factual assumptions on which he chooses to base his deductive arguments, provided his empirical conclusions seem to be correct. The prevalence of such a point of view is, to a large extent, responsible for the state of splendid isolation in which our discipline nowadays finds itself. Leontief (1971, p. 5)

⁸Kuznets gave real encouragement to many outstanding economists to follow his sound methodological approach. For example, he was the mentor of Moses Abramovitz who in turn strongly influenced Paul A. David.

⁹A major or radical or breakthrough innovation is one that establishes a new platform for the launching of many subsidiary innovations and improvements. For definitional issues about the notion of major technological innovation see Kuznets (1973, esp. pp. 190–196). More on this in Section 6.

¹⁰According to Schumpeter (1955, p. 260) an equivalent statement of the law of diminishing returns was first formulated by Turgot and ‘embodies an achievement that it is nothing short of brilliant and suffices in itself to place Turgot as a theorist high above A. Smith.’

¹¹It goes without saying that the literature on evolutionary economics is too extensive to cite.

¹²The analysis of statistical data induced N. Kondratiev to conclude in the early 1920s the existence of long cycles with an average length of 50 years. These long cycles or long waves are usually referred to as ‘Kondratievs’.

¹³Analysts at the Institute for the Future have observed that interactions between cutting-edge fields of knowledge can evolve in unexpected directions.

¹⁴Additional aspects on the limitations of the Kuznets can be found in Section 6 of the present paper.

¹⁵The following examples (radio, computers, and laser) are taken from Rosenberg (1995).