Racing or spilling? : the determinants of research productivity in ethical drug discovery

This paper presents the results of a study of the determinants of research productivity in the pharmaceutical industry. Using disaggregated, internalfirm data at the research program level from ten major pharmaceutical companies, we find no evidence of increasing returns to scale at either the firm or the research program level. However our results suggest that there are three benefits to running research programs within the context of larger and more diversified R&D efforts: economies of scale arising from sharing fixed costs; economies of scope arising from the opportunity to exploit knowledge across program boundaries within the firm; and the enhanced ability to absorb internal and external spillovers. We alsofind that spillovers between firms may play a major role in increasing research productivity. The paper also speaks directly to the question offirm heterogeneity. A significant proportion of the 'firm effect" identified in previous studies can be explained by the slowly changing composition of the research portfolio, as well as by less easily measured aspects of innovative capability. Introduction While there is considerable agreement that economies of scale and scope in performing R&D and the magnitude of R&D spillovers may have important implications for research productivity, research to date has been far from conclusive. The theoretical literature is divided, for example, as to the effects that relative firm size or the presence of spillovers should have on incentives to undertake research (Dasgupta and Stiglitz (1980); Spence, (1984)), while empirical work has generated surprisingly inconclusive and sometimes contradictory results. (Cohen and Levin, (1989).) Several observers have suggested that this is the result of a historical reliance upon aggregated data, combined with a failure to control adequately for firm or industry effects (Baldwin and Scott, (1987); Cohen and Levin, (1989)). In this paper we look inside the firm for evidence on the importance of scale, scope and spillovers for research productivity, using detailed, disaggregated, data at the level of the research program obtained from the internal records of ten major research-oriented pharmaceutical companies. The enriched understanding of the production technology of innovation which this research provides may help to resolve several debates in the economics of industrial organization. There are also implications for the theory of the firm. For example several scholars, notably Chandler (1990), have suggested that the presence of very large firms in modern capitalist economies is driven by the