Little cause for concern: Analysis of gender effects in structured employment references

Abstract

Recent research has highlighted the fact that narrative letters of recommendation in employment references could contribute to gender bias in personnel selection. Structured, quantitative employment references, however, may limit the opportunity for such biases to emerge. In a sample of nearly one million applicants and ratings by over four million employment reference providers, we found no meaningful effect of gender bias in highly structured, quantitative employment references across job levels and a wide variety of industries. Interestingly, and in contrast to existing theory, the effect of gender bias remained negligible across both stereotypically masculine and feminine jobs. Similarly, in a subsample of 5000 job applicants and 20,000 employment reference providers, coded verbatim comments of reference providers showed little practical gender differences in the frequency with which various comment types are made. These results suggest that highly structured, quantitative and semi-structured, verbatim employment references are an effective tool in the advancement of fair and equitable personnel selection practices. Theoretical and practical implications are discussed, and future research is proposed.