BETA's De-Identified Recruitment Trial — Anonymised Shortlisting Did Not Help Women in the Australian Public Service (Cautionary Case)

In 2017 the Australian Government's Behavioural Economics Team (BETA) published 'Going blind to see more clearly', a randomised trial of de-identified recruitment at the shortlisting stage.
Over 2,100 public servants from 14 Australian Public Service agencies shortlisted applicants for a hypothetical senior role; participants were randomly assigned to see standard or de-identified CVs. Reviewers were 2.9% more likely to shortlist female candidates and 3.2% less likely to shortlist male candidates when identities were visible, and minority candidates were also shortlisted more often when identifiable. Behaviour varied considerably across agencies and reviewer groups.
BETA concluded that de-identifying applications at shortlisting does not appear to promote diversity and may frustrate it, reading the pattern as a mild form of affirmative action by reviewers. It suggested that attention might be better directed to other recruitment stages. Limitations: the exercise used hypothetical candidates and a hypothetical role, so real-world hiring outcomes were not measured. The case is included as honest evidence that a popular, well-intended measure can fail.

Read the full analysis: https://www.pmc.gov.au/sites/default/files/2025-06/unconscious-bias.pdf