Selecting technological paradigms beyond push-pull dynamics

Giovanni Dosi\´s technological paradigm theory was developed in part to correct the then dominant practice to focus on either "demand-pull" or "technology push" when explaining technical change (Dosi, 1982). In Dosi\´s view demand is not, as demand-pull theorists propose, the first motivation for technical change. However, he also considers that pure technology-push models failed to recognise the obvious importance of economic factors. He proposes two roles for demand: (i) demand influencing selection among competing paradigms; and (ii) demand influencing the course of the paradigm after its inception. The literature on the paradigm framework mainly focuses on the second of these selection mechanisms, i.e. the role of the market environment in shaping the precise trajectories of advancement within the set allowed by a specific paradigm. Instead, the mechanism we study in this paper has received little attention; it describes the role of demand in the selection by the supply side amongst technological paradigms upstream in the innovation process. One reason for this lacuna in the literature could be the difficulty involved in observing paradigms before their materialisation into marketable devices. An important part of the \´variation\´ (in the spirit of the evolutionary theory) on which this pre-market selection process acts is based on ideas; selection in terms of the market environment is more easily observable, as it is based on commercialised devices. Observing variation that exists only in the human mind is a common problem in the broader field of evolutionary economics and even in general evolutionary approaches to social sciences (Nelson, 2006). However, if we consider the technological \´pre-practice variation\´ to be not only the ideas in engineers minds, but also (somewhere downstream of the purely mental realm, but largely upstream of market environment selection) the observable activities of research and experimentation, we can learn more about the empir- ical functioning of this pre-market selection dynamics. This article will try to show this selection mechanism in operation by means of a case study of the technological evolution of the artificial disc, an orthopaedic prosthesis employed in the treatment of chronic back pain. We will consider how the supply side selects between two technological paradigms of this prosthesis, not in the form of plausible designs in engineers\´ minds, but in the form of research projects, whose development to become marketable innovations depends on this pre-market selection process, expressed in the success or failure of these projects to become marketable products. To understand the role of demand in this pre-market selection mechanism we develop a conceptual framework which includes the relationship between sales, R&D investment and the intrinsic technological resistance of two artificial disc paradigms. The pre-market selection process of paradigms will occur after an industry learning process \´discover\´ whether the R&D project investment derived from sales conditions is enough to exceed the technological resistance of the paradigms. We illustrate how this learning is achieved through various artificial disc research projects, helping to explain the historical supply side choices in the evolution of the artificial disc: our main results show that changes in sales conditions led to the R&D investment reconsideration of an artificial disc paradigm that had been abandoned before because of the excessive effort needed to exceed its intrinsic technological resistance under the previous sales conditions.