Gendered language in academic evaluations : evidence from the Italian university system

We analyze the impact of evaluator and candidate gender on the language used in academic evaluations using data on 295,000 evaluation reports for applicants seeking professorial promotion across all academic fields in Italy. In this context, candidates are assessed by a national-level committee composed of five randomly selected evaluators from the corresponding field. We observe that the language used in evaluation reports varies significantly with applicants' productivity and professional ties to evaluators, but we find no indication that the language of the assessments depends on the gender of either the candidates or the evaluators.