Does gender really matter? : an analysis of Jena University scientists collaboration with industry and non-profit-partners

One of the dominant changes in academics during the last 25 years has been the steadily growing political pressure on universities to strengthen their links with and knowledge transfers to external partners. This focus on university-industry collaboration has been accompanied by another fundamental change in academia, the growth in enrolments of female researchers in science (especially natural sciences) and engineering. The rising number of female scientists has led to the question of possible gender disparities as to external collaboration activities, namely a male predominance as to external collaboration activities. This paper extends the existing empirical work in three respects. First, it covers various scientific disciplines and faculties, second different types of transfer activities and third different transfer partners, including private industry, the civil service and non-profit-organizations. Data were collected from a sample of university professors at two German universities. The resulting survey is based on 174 personal interviews lasting 40-60 minutes. The empirical results point to a more complicated story than gender differences of productivity or simple discrimination. As to collaboration activities of female professors with the business sector the empirical evidence is not uniform and straightforward. With our data set it is easy to produce empirical outcomes seemingly confirming the gender gap, e.g. by omitting variables as to scientific fields like engineering or business economics. Even including all relevant control variables the results are influenced by the specification of the empirical model: The level-level-model strongly corroborating and the log-log-model refusing a significant negative difference related to gender. But as to the collaboration with the public sector and non-profit organisations there is no gender gap at all and this outcome does not depend on the specification of the regression equation. Thus, specific differences of collaboration partners seem to play a role.