Change agents and the evolution of the United States pharmaceutical industry, 1920--1960

This dissertation examines the evolution of the market and the firms in the context of an emerging innovative race in the U.S. pharmaceutical industry. The history shows the existence of two distinct industry regimes. Before 1940, the industry can best be characterized as a competitive regime in which a strategy based on cutting prices was viable. By the 1950s, a strategy of seeking innovations became visible and dominant in the industry. This historical observation raises the following questions: What made this regime change? Were there change agents who made creative responses, in Schumpeter\u27s term? The answers are sought in the theoretical framework of Nelson and Winter\u27s evolutionary economics. It is argued, in this dissertation, that important changes in markets typically occur in two stages. First, innovators or change agents foresee new business opportunities through routine search behavior (i.e., R&D activities) and mobilize scarce firm resources in search of new ways of surviving (variation mechanism). Second, when this innovative search is successful, change agents grow faster than the other normal agents and become a threat to them (selection mechanism). Theoretically, these actions break down an existing economic order. In this picture, business organizations with routine search behavior occupy a central place in the unfolding of economic change. The change in industry state is conceived of as the joint outcome of change agents\u27 actions ( creative response ) and normal agents\u27 reactions (Winter, 1971; Nelson and Winter, 1982). The objective of this dissertation is to determine whether this two-stage mechanism of change has an empirical counterpart in the context mentioned above. The analyses of SD regressions demonstrate the operation of variation mechanism in the observed history. That is, those firms which were able to generate new antibiotics, (change agents) revealed remarkably different search behavior than those which were not (normal agents). These analyses indicate approximate symmetry of search behavior between these two agents until the late 1930s, but after 1940, they clearly demonstrate a significant asymmetry. The change agents identified in this study systematically hired more biologists, as well as other scientific staff prior to the generation of new antibiotics, which were the best-selling wonder drugs in the late 1940s and 1950s. This study shows that the mechanism of selection operated in the period roughly from 1950 to 1960. SD regression models clearly indicate that the identified change agents grew much faster than the average firm in the market. Poisson regressions demonstrate that these firms also did significantly better in generating other new chemical entities. Interestingly, a majority of pharmaceutical firms failed to adapt to the change and remained marginal players in the market. Additional secondary information shows that these winners dominated the market in the two decades following 1960. In sum, this study corroborates the central part of evolutionary theory in the midst of an emerging innovative race in the U.S. pharmaceutical industry