The Effect of Article Length on the Share of Citations of Women in Articles

• In Event: New Methods for Text-as-Data

Abstract

Citations of published articles are often used as a measure for ‘impact’ and as a metric in academic hiring and promotion processes. The important role of citations make the presence of a gender citation gap, men being cited more often than women, highly problematic. It reduces the visibility of work by women and disadvantages them in the pursuit of an academic career.
Research on gender and citations produced evidence that there is a gender gap and how it varies across subfields and depends on author-specific and field-specific characteristics. In comparison, factors that are related to characteristics of journals and their publishing requirements have received less attention. It is important to study the effect of journal characteristics and editorial requirements on the gender citation gap. If it was found that such factors have an effect on how often men and women are cited, then one could change journal and editorial requirements such that the citation gap is reduced.
In our paper, we estimate the effect of article length on the share of citations of work by women in an article. We hypothesize that shorter articles have a lower share of citations of women compared to longer articles. The mechanism is that shorter papers have to cite fewer articles and account for more selective citation behavior. Higher selectivity leads to disproportionately more citations of work by men. First, one is more likely to cite authors from their own academic network, which largely includes other men when a man is an author. Second, selectivity means that one cites more prominent, widely known research from the past that has been mainly written by men because academia has been largely male-dominated then.
We estimate the effect of article length by comparing short formats (letters, research notes etc.) from political science journals to traditional, long-form articles. We compare the share of citations of women-only and men-only publications between short and long formats of the same journal. An intra-journal comparison of formats allows us to control for journal-specific factors that might confound the gender-citation relationship. A second, complementary identification strategy that we use is text matching to pair short and long articles that have comparable content. The empirical results of our analysis will show whether short formats, which allow researchers to study more focused research questions and to publish faster, come at the expense of contributing to a gender citation gap in political science journals.

Authors

• Barbara Ellynes Zucchi Nobre Silva, Universität Passau

• Ingo Rohlfing, Universität Passau

• Esther Song, University of Bergen