Universality of citation distributions for academic institutions and journals

TL;DR

This study reveals that citation distributions for individual institutions and journals, when scaled by their averages, follow a universal pattern, indicating common underlying citation dynamics across diverse academic entities.

Contribution

It demonstrates the universality of citation distribution shapes across institutions and journals after rescaling, highlighting a common pattern in citation inequality.

Findings

Citation distributions can be rescaled to a universal form.

High inequality in citation distribution with Gini coefficients around 0.66 and 0.58.

Top 25% of articles account for about 75% of citations.

Abstract

Citations measure the importance of a publication, and may serve as a proxy for its popularity and quality of its contents. Here we study the distributions of citations to publications from individual academic institutions for a single year. The average number of citations have large variations between different institutions across the world, but the probability distributions of citations for individual institutions can be rescaled to a common form by scaling the citations by the average number of citations for that institution. We find this feature seem to be universal for a broad selection of institutions irrespective of the average number of citations per article. A similar analysis for citations to publications in a particular journal in a single year reveals similar results. We find high absolute inequality for both these sets, Gini coefficients being around $0.66$ and $0.58$ for institutions and journals respectively. We also find that the top $25$% of the articles hold about $75$% of the total citations for institutions and the top $29$% of the articles hold about $71$% of the total citations for journals.