Three Essays on the Development and

The thesis comprises three essays on various aspects of the development and diffusion of pharmaceutical innovations, woven together by the idea that the production of clinical knowledge inßuences organizational design, pricing, and advertising decisions in the pharmaceutical industry – that it is an essential ingredient of pharmaceutical Þrms’ technology strategy. The Þrst essay studies why pharmaceutical Þrms partly contract out the operational aspects of clinical trials to Contract Research Organizations. Using detailed project-level data, I Þnd that (1) outsourcing intensity is sensitive to surprises in the demand for clinical trials services; (2) data-intensive projects are more likely to be outsourced than knowledge-intensive projects; (3) Þrms that consistently increased their share of outsourced activity over time had more productive internal teams than those that did not. This last result draws attention to the perils of considering a single transaction as the unit of analysis when explaining shifts in Þrm boundaries. The second essay investigates how different sources of information inßuence the diffusion of new pharmaceutical products. Using a novel index of clinical-research output, I Þnd that both marketing and published clinical results directly inßuence the diffusion process in the anti-ulcer drug market, but scientiÞc outputs do not seem to be important drivers of Þrms’ marketing efforts. The direct effect of science on demand implies strong private incentives for clinical research. In the third essay, Ernst Berndt, Robert Pindyck and I examine the role of consumption externalities in the demand for pharmaceuticals. These effects emerge when use of a drug by others affects its value or conveys information about safety and efficacy to patients and physicians. This can affect the rate of market diffusion for a new entrant, and lead to herd behavior whereby a particular drug can dominate the market despite the availability of close substitutes. We use data for anti-ulcer drugs to estimate a dynamic demand model and quantify these effects. We Þnd that consumption externalities inßuence both valuations and rates of diffusion, but that they operate at the brandand not the therapeutic-class level.