Recruiters spend less than 10 seconds screening a CV. Such rapid decision-making increases the role that bias can play in hiring decisions. Rather than skills, irrelevant factors can become filtering criteria or sources of discrimination, consciously or otherwise.
To test how changes to the design of CVs might reduce a bias against gaps in employment, we ran a trial sending over 9,000 CVs to real job vacancies.
Women returning to work after caring may face an unlevel playing field
Having a gap on your CV makes recruiters more likely to dismiss an application, affecting those returning to work after taking time out for caring responsibilities. The vast majority (91%) of ‘returners’ are women, so addressing this problem is important for improving workplace gender equality.
We wanted to understand whether small changes to the way CVs are presented would increase positive responses from employers.
We first tested whether it is better to leave an employment gap unexplained or explain that it was for childcare. We suspected that an explanation would increase bias. Experimental US evidence finds that gaps for childcare are penalised more than gaps for unemployment. However, when we interviewed UK HR professionals, they felt it was important to explain gaps taken for care. We wanted to understand whether what people say aligns with what they do.
It’s a common finding in behavioural research that people don’t always act in the way they think they will, so we also tested whether removing dates of experience and instead presenting previous roles in terms of the number of years experience removed the opportunity for bias against a gap.
We applied to more than 9,000 real-world job vacancies
We ran a randomised controlled trial comparing four equivalent CVs and cover letters, with the following alterations:
- Currently employed: No gap
- Explained gap: 2.5-year gap since last job explained for childcare
- Unexplained gap: 2.5-year gap since last job left unexplained
- No dates: Dates of employment history replaced with the number of years’ experience
We applied to over 9,000 job vacancies over a 6-month period October 2019 to March 2020 in a combination of higher and lower skilled roles in male- and female-dominated industries. We recorded the rate of positive callbacks, e.g. an invitation to interview, a job offer or progress to the next stage of the application.
Removing dates increased positive callbacks
Importantly, we did not find a difference in callback rates whether the gap was explained or left unexplained, despite what HR professionals told us. For lower skilled roles, the CV explaining the gap performed the worst, which is a concern since returners are more likely to have lower levels of formal education on average.
Presenting previous roles in terms of the number of years of experience rather than the dates of that experience increased positive callbacks by 14.6%. We speculate that this format encourages employers to see applicants as experienced where dates may activate irrelevant associations or stereotypes.
The 2 percentage point difference in callback rates for CVs with and without gaps was not statistically significant. However, this finding should be interpreted with caution since other research finds discrimination against a gap, suggesting this may have been due to specific contextual factors of the trial.
We suggest changing guidance and default CV settings
These findings take on even greater relevance during this period of COVID-induced economic uncertainty, where vast unemployment may increase the number of people with significant CV gaps.
Encouragingly, the results have already led the government to plan a review of their guidance, which previously encouraged returners to provide details about their gap for care.
Alongside guidance, small changes to the way job sites or similar websites format CVs and employment histories by default could have a big impact in supporting people back into work. In particular, we suggest removing any requirement to enter dates and providing the option to display experience in terms of years.
The trial was part of the Gender and Behavioural Insights (GABI) research programme, funded by and in partnership with the Government Equalities Office, who approved the trial design. This research is in the process of being submitted to a journal for full peer review and publication.