One Point of View: Corporations Need to Publish — or Perish

Abstract

A number of tensions exist between the R&D departments and the business managers in most high-technology companies today. A principal one centers on finance: How much money will be spent on R&D? A second involves the balance between basic and applied research, as scientists generally seek more basic while the business managers prefer more applied. A third tension revolves around deciding on the appropriate metrics of R&D output (papers, patents, products, and revenues). A potential source of conflict that has received somewhat less attention than these three is the general desire among research scientists to publish their findings as quickly as possible. Business managers, not surprisingly, fear lost secrets and the corresponding profits, and thus rarely encourage publication. Our purpose in this article is to enumerate the reasons why managers should support their scientists' publication efforts. We should make clear that we recommend this policy of openness not because we are academic purists, but for practical profit-focused business reasons. A Little Sociology From the very beginning of their doctoral studies, research scientists are conditioned to believe that publication is important, whether they choose to work in academia or in industry. Most of the theoretical background for this originates with the sociology of science literature and Robert Merton (1). Merton argued that a significant factor in the advance of science stems from the idea of priority. The first person to announce a discovery derives multiple benefits from that discovery. These benefits can be economic in nature, but there may also be social and reputational effects. Priority of discovery has long been recognized as the means by which someone builds a reputation in the scientific community, and one's reputation for contributions to the field is a principal in science, particularly for academic scientists. This currency can increase a scientist's ability to change employers, be awarded grants, and to have access to the best equipment and personnel. Thus, priority is a fundamental facet of the reward system, and it sets up a continuous race to be first. However, to achieve priority, scientists must tell the world about their findings (i.e., publish), and it is in this area that conflict with industrial management may arise. Why Companies Publish Previous research has found that many companies are reluctant to publish the findings of their research scientists because publication might disseminate information that could be to the firms' economic disadvantage (2). For example, Larry Ellison, CEO of Oracle, reportedly got his initial idea about relational databases from a scientific article published by an IBM researcher, Ted Codd (3). We suspect that IBM's legal department would consider that publication to have been premature at the least! However, other research has found that industrial researchers are prolific in their publication efforts, and that these efforts can generate high-quality research (4,5). Also, in another research effort, we studied 20 pharmaceutical companies over a 13-year period. We found that firms with Ph.D.-trained CEOs and other top managers were more likely to encourage publications by their scientists and, more importantly, that open firms were much more successful in their R&D efforts, as measured by both new drugs developed and future economic performance (6). These findings regarding the value of openness may appear counter-intuitive, yet further examination suggests that such openness is both logical and appropriate. Even though there are a few instances, like Oracle, where publication may have been detrimental to a firm, they are actually quite rare. The primary reason is that it is very difficult to replicate an experiment or new technology based solely on a reading of published accounts. Collins and colleagues investigated the question of replication by assessing whether researchers who were attempting to construct a new type of laser could do so by reading scientific articles on the subject (7). …