Abstract

Objective: The aim of the research is to assess the research and development efficiency of the institutes of the Polish Academy of Sciences (PAS) in 2019 and to identify external factors that have a significant impact on it.
Research Design & Methods: A two-stage procedure in the field of DEA methodology was used. In the first stage, the efficiency of PAS institutes was estimated using alternative BCC and SBM models. In the second stage, a Tobit model was used to isolate external factors significantly influencing efficiency. Given the data available, two types of research and development (R&D) efficiency were analysed: publishing efficiency and combined publishing and implementation efficiency.
Findings: A significant share of institutes are highly inefficient (nearly a half of the units in terms of publishing efficiency and a third in terms of combined efficiency). The fields in which a given scientific unit conducts research which significantly affect publication efficiency differ from those that significantly affect combined efficiency. Both types of efficiency are significantly negatively affected by the increase in the scientific category of the unit.
Implications / Recommendations: The source of high research and development inefficiency among a significant number of institutes is the fact that they also generate other outputs than those considered in the work. The majority of institutes do not apply to the R&D sphere. The significant negative impact of the increase in the scientific category on R&D efficiency indicates that future focus should prioritise the quality of publications over their quantity.
Contribution: The efficiency of the research and development activities of the PAS institutes was assessed after the introduction of the last reform of Poland’s system of science and higher education in 2018 (such studies have yet to be carried out). It is also important to use a two-stage approach within the DEA methodology in order to isolate external factors that significantly
influence this efficiency.

Citations

Authors (2)