AN ANALYSIS OF THE SOURCE AND NATURE OF TECHNICAL CHANGE: THE CASE OF U.S. AGRICULTURE Jean-Paul Chavas, Michael Aliber, and Thomas L. Cox* Abstract —This paper proposes a methodology to investigate the process of technical change with a focus on the dynamic effects of R&D investments on productivity, and on the induced innovation hypothesis for both inputs and outputs. The approach builds on a nonparametric representation of the underlying technology. An application to U.S. agriculture is presented. By distinguishing between private and public R&D investments, the analysis provides useful insights into the source and the dynamic nature of technical progress. I. Introduction M UCH research has been done on technical change and its economic significance. The importance of technical change in economic growth is widely acknowledged (e.g., Solow (1957)). Yet the process of technical change is quite complex and still poorly understood. It is often described as a dynamic two-stage process: (1) the creation of new knowledge and technology, and (2) the adoption of new technology by firms. For example, in his study of an innovation cycle, Griliches (1957) identified significant lags between the creation of new knowledge and the adoption of the associated new technology. The creation of new knowl- edge is stimulated by expenditures on research and develop- ment (R&D). Because of the cost of acquiring new knowl- edge and adapting it to local conditions, the adoption process for a new technology can be slow. The lags between R&D investment and its payoff can vary with each technology and each industry. However, as a rule, these lags are longer as the research is more basic, and shorter as the research is more applied. A distinction often made in the literature concerns the source of funding for R&D, namely, public versus private. When new knowledge has the characteristics of a public good, then public funding of research may be appropriate. Alternatively, when property rights to technology can be privately assigned and enforced (e.g., through patents), then private institutions can have the proper incentive to invest in R&D. But because patent rights expire at the end of the patent life, private incentives to invest in basic research with longer term (and more uncertain) payoffs are weak. As a result, basic research tends to be funded publicly, and private R&D tends to focus on applied research with shorter term payoffs that can be appropriated during a patent life. The linkage between technical change and resource scarcity has been the subject of much scrutiny. By definition, technical progress allows the production of greater outputs with the same amount of resources, or the use of fewer resources to produce the same outputs. Thus it can help reduce resource scarcity. The feedback effect of resource scarcity on technical change is also of interest. It has been expressed in terms of ‘‘induced innovation.’’ As first sug- gested by Hicks, the induced innovation hypothesis states that relative resource scarcity tends to guide the process of technical change (Hicks (1932)): A change in relative prices of the factors of production is itself a sign to invention, and to invention of a particular kind—directed to economizing the use of a factor which has become relatively expensive. The induced innovation hypothesis is typically formulated in terms of the bias in technical change (e.g., see Binswanger and Ruttan (1978)). Roughly stated, technical change is said to be biased toward a particular factor (or factor using) if it stimulates the relative use of this factor. Conversely, techni- cal change is biased against a particular factor (or factor saving) if it reduces the use of this factor relative to other factors. In this context, the induced innovation hypothesis predicts that technical change will be biased against a particular factor (i.e., factor saving) when this factor’s relative scarcity (e.g., its relative price) increases. Con- versely, technical change will be biased in favor of a given factor (i.e., factor using) when its relative price declines. The induced-innovation hypothesis has been subject to empirical testing. Binswanger (1974) and Hayami and Ruttan (1985) found empirical evidence in support of the hypothesis, whereas Olmstead and Rhodes (1993) uncov- ered some historical evidence inconsistent with induced innovation. This research has typically measured the bias in technical change and compared its direction with changes in relative prices. The linkage between this bias and the nature of invention has often not been made explicit. This is unfortunate, since Hicks’s original formulation of the in- duced-innovation hypothesis explicitly mentions inventive activities. This suggests a need to explore jointly the economics of R&D investments and induced innovations. Given the dynamics of R&D effects, this indicates that induced innovations should also be investigated in a dy- namic setting. The objective of this paper is to analyze the process of technical change with a focus on the effects of R&D investments on productivity, and on the induced-innovation hypothesis. This is done relying on nonparametric methods developed by Afriat (1972), Hanoch and Rothschild (1972), and Varian (1984) (see section II). The nonparametric approach to production analysis consists in analyzing a finite body of data without ad hoc specification of functional forms for the production function, supply–demand func- Received for publication January 30, 1995. Revision accepted for publication April 11, 1996. * University of Wisconsin-Madison. We would like to thank two anonymous reviewers for useful comments on an earlier draft of the paper. This research was supported in part by a Hatch grant from the College of Agricultural and Life Sciences, University of Wisconsin-Madison. 3 482 4 r 1997 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/003465300556896 by guest on 14 September 2021