Thesis

Kwanghui Lim, MIT Sloan School

Abstract

This thesis comprises three essays on the relationships among basic research, applied research, and innovation. Earlier research emphasized that absorbing external knowledge requires effort and investment (Cohen and Levinthal, 1989; 1990). This thesis explores various mechanisms through which absorptive capacity is developed, including a firm's R&D, its connectedness to the external scientific community, the provision of a science-oriented research environment, and investment in basic research. The chief contribution of this dissertation is to document the many ways in which firms develop absorptive capacity, and how absorptive capacity varies across industry, stage of technology development, and scientific area.

The first essay explores how firms develop different kinds of absorptive capacity. A firm’s absorptive capacity depends upon internal R&D and its connectedness to universities, other firms and R&D consortia. R&D is effective for absorbing disciplinary knowledge; alternative mechanisms are useful for domain-specific knowledge. A science-oriented research environment is not necessary, as long as the firm remains connected through other means. To illustrate, I trace knowledge spillovers of copper interconnect technology for semiconductors.

The second essay examines the concentration of basic and applied research relative to innovation. In the semiconductor industry, basic research is surprisingly concentrated relative to innovation. Since spillovers are prevalent in this industry, I conclude that many semiconductor firms capture spillovers without performing much basic research. In the pharmaceutical industry, basic research and innovation have similar concentrations. In both industries, applied research is not concentrated relative to innovation.

The third essay examines researchers at five firms. Given two researchers with the same number of publications at IBM, AT&T, or Intel, the one who publishes a higher fraction of her papers in basic research journals is less likely to patent. These researchers face a tradeoff between participating in basic and applied research (Allen, 1977). The opposite holds at Merck and DuPont, where researchers who publish a higher fraction of papers in basic scientific journals obtain more patents. Thus, basic research has a positive impact on pharmaceutical patents (Gambardella, 1992; Cockburn and Henderson, 1998). Within Dupont and Merck, patenting is most closely associated with publications in basic chemistry, and with pharmaceutical R&D.

Thesis Committee: Scott Stern (Chair) Assistant Professor Alfred P. Sloan School of Management, MIT Rebecca Henderson Eastman Kodak LFM Professor of Management Alfred P. Sloan School of Management, MIT Eric von Hippel Professor, Management of Innovation Alfred P. Sloan School of Management, MIT