The Organization of Innovation: Theory and Evidence from the Pharmaceutical Industry
Abstract: A firm that develops a new product potentially cannibalizes sales of existing products in the firm's product portfolio, where such cannibalization is more costly the more profitable are sales of the cannibalized products. Thus, if the firm is currently producing a product for which its market power is substantial, it will want to control the research and development process in order to limit cannibalization. In this paper, we explore how this basic logic affects the organization of investments in research and development. We first build and analyze a theoretical model of the research and development process in which conducting R&D in-house provides the firm more control over the new product's location in product space. We then explore the model's testable predictions using data from the pharmaceutical industry concerning patents, patent expiration, and decisions concerning whether various stages of the research and development process are conducted in-house or outsourced. Our empirical findings support the model's testable predictions.
