Estimating the Licensing Probabilities in the Academic Context: An Empirical Analysis

Abstract

Licensing technologies are one of the main ways to produce and bring academic research to society. Despite previous studies’ dedicated efforts to identify licensing probabilities, the question of how the expertise and prestige that a university has in a given technological field influences the licensing probabilities is still little addressed. This article aims to identify information in patent documents to estimate the probabilities of licensing technologies produced at the university. For that, we performed a data mining of licensed and unlicensed patents from an important Brazilian University (n = 1,578). We estimated the licensing probabilities using the Logistic Regression technique, based on the Maximum Likelihood Estimation. The results suggest that the variables of know-how in the main field and Technological strength in the main field are the most important/influential variables in estimating the probabilities of licensing a given patent. The main conclusion obtained from the results is that: universities, to obtain more licenses, must increase their know-how (expertise) in some technological fields, maintaining a reasonable level between specialization and diversification. Additionally, the higher the citations received (prestige/recognition) by a university in a given technological field, the greater the probability of patent licensing in that technological field. In terms of practical contributions, this study suggests that: investments in specific technological fields generate more competitive advantages for the university and, thus, more technological successes.