Subject mix and productivity in Portuguese universities

Recent changes in higher education, especially associated with its massification, have led to calls for more accountability and to an increasing need to evaluate universities' performance. However, comparing universities with different subject mixes poses some significant problems. In this article, a linear mixed effect model methodology allowing for random effects at university level is used to measure the productivity of Portuguese public universities, taking into account their subject composition, for a period of five years (1997–2001). The results show that subject mix is an important factor in explaining the variability in productivity between universities. However, even taking account of subject mix, a lot of inefficiency is found in the system, and a significant variability in productivity among universities exists.