On evaluating author's performance by publications: an axiomatic study

Administrators in all academic institutions across the world have to deal with the unenviable task of comparing researchers on the basis of their academic contributions. Unfortunately, however, there is no reasonably established consensus on the method of arriving at such comparisons, which typically involve trading off accomplishments in teaching, grant writing and academic publication. In this paper, we focus on the particular dimension of academic publication, and analyze this issue from a more fundamental perspective than addressed by the popular h-index (which may lead to unfair and counter-intuitive comparisons in certain situations). In particular, we undertake an axiomatic analysis of all possible ways to measure academic authorship for a given dataset of research articles and find that an egalitarian eindex is the only method which satisfies the axioms of anonymity, monotonicity, and efficiency. This index divides authorship of joint projects equally and sums across all publications of an author.