- Over half of job ads in the UK contain more masculine phrases than feminine
- 85% from ten sectors contained at least one ‘masculine phrase’
- Only 16% of job adverts have an even balance of gendered wording
- Adverts in Media & Marketing, Financial Services and Tech sectors most commonly reference gendered wording
New research released today shines a light on systemic bias within the UK job market towards male applicants. The findings show that over half (52%) of job adverts are phrased with gendered wording which makes them unconsciously biased towards men. The study from global recruitment specialists Michael Page and the University of Manitoba, highlights how female candidates are being subtly deterred from applying for roles, across several sectors, even before submitting an application.
The findings are concerning for women, many of whom may have already experienced a disproportionate financial impact directly resulting from the pandemic and subsequent recession. Recent findings have shown that women globally have been 1.8 times more likely to lose a job due to COVID-19 than men. With this in mind, subtle gender biases within the workplace, such as gendered language, are likely to hold back the rate at which women can get back on the job ladder.
The expert team who explored the gendered language bias included social psychologists with experience of analysing gendered wording in job adverts and Michael Page consultants. They reviewed over 5,000 job adverts from ten different industries issued by five of the UK’s leading recruitment companies. Following their analysis, the team established a consistent presence (85%) of phrases with masculine connotations, such as ‘leading’ and ‘analytical’. The presence of which is likely to be contributing towards women being unconsciously dissuaded from applying for certain jobs.
Media & marketing, financial services and tech were shown to be the sectors with the largest disparity between the usage of masculine and feminine words. Contrastingly some sectors, such as HR, that make a conscious effort to drive inclusivity had a closer parity between the number of masculine and feminine phrases used in their adverts.
The study also established the most commonly used gendered words – as categorised by existing gendered language literature across ten sectors, which included:
- Media & marketing
- Financial services
- Leisure & tourism
- Manufacturing & industrial
- Public sector
Variations of the phrase ‘lead’ were determined to be the most commonly used ‘masculine phrase’ across all sectors except for financial services and retail sectors, where ‘analyse’ was the most common.
All ten sectors listed variations of ‘support’ as the most commonly used feminine word. When taking account of all ten sectors, the researchers assessed that just 16% of all job adverts comprised an even balance of feminine and masculine gendered word.
Commenting on the results Sheri Hughes, Director of Diversity and Inclusion at Michael Page, said:
“The process of being more inclusive involves both listening and reacting to important developments, such as the findings our research has uncovered today. It’s accepted that women have been disproportionately impacted by the pandemic, with many struggling to get back to work after being made redundant. We believe that everyone deserves an equal opportunity to bounce back from the effects of COVID-19, regardless of their gender.
“Our findings highlight there is still significant work needed to break many of the conscious, as well as unconscious, institutional barriers which prevent women from achieving equality in the workplace.
“We’re urging companies across all industries to take these findings seriously and to join us in committing to raising awareness of gender bias within job adverts and encourage other recruitment firms to join the fight for a fairer job market for all. The more inclusive the language the better the attraction for candidates across all under-represented groups. This will be beneficial to businesses in the long term, as they maximise the chances of finding the right talent for the roles they have on offer.”
Michael Page is raising awareness of gendered wording in job adverts, with the ambition of making it balanced across the industry. The company provides software tools to its consultants, which can identify and recommend ways of phrasing sentences to ensure the language used remains neutral. In addition, the leadership team undergo mandatory unconscious bias training and have measurable D&I objectives, which directly impact their remuneration.
Danielle Gaucher, Associate Professor of Psychology and Director of the Intergroup Relations and Social Justice Laboratory at University of Manitoba, said whilst reviewing the findings:
“There are people who want to believe that these types of subtle biases don’t still exist in 2021, but they are still prevalent across all areas of society, including hiring and recruitment. It’s important not to underestimate the power, and the insidious impact that certain words can have towards our thoughts and feelings.
“Our findings demonstrate how powerful words can be when applied to the recruitment environment. It is also important to distinguish that gendered wording does not imply a person of the opposite gender cannot exemplify a particular trait, such as men being supportive and women being analytical.
“However, many do not recognise the implications that gendered wording can have, especially in deterring women for applying for jobs, and how it can act as one barrier to improving gender equality in the workplace. We have tools now to identify problematic job adverts and to help solve this equity issue.”
Crystal Akass, Director of People Services, People and Capability Group, Department for Work and Pensions, commented:
“The Department for Work and Pensions welcomes organisations like Michael Page shining a light on this issue and was pleased to partner with them on a recent campaign to attract and recruit a cadre of diverse and talented HR professionals.”
The study was carried out by randomly selecting 5,000 job adverts from five leading UK recruitment firms. The job adverts, which were live between February and March 2020 spanned across ten different employment sectors.
The adverts were then examined using Linguistic Inquiry and Word Count software (LIWC) which coded them for the presence of pre-established gendered wording which had been identified through gendered stereotyping literature reviews. The expert team also built a specialist software programme which was able to generate the most common and unique words listed across all the job adverts, as well as categorise them by sector. All statistical analysis was conducted using IBM SPSS software.
Michael Page also have a Gender Decoder tool which enables users to quickly check job advertisements for any specific words which may appeal to one gender over the other. For more information on gendered wording in job adverts within the UK, please contact the below press contacts.
5,000 job advertisements were sourced from five UK recruiting companies (Michael Page, Hays, Indeed, Reed and Total Jobs) during February 2020. These job advertisements spanned across 10 jobs sectors, including: Business Services, Financial Services, HR, Legal, Leisure & Tourism, Retail, Manufacturing & Industrial, Public Sector, Media & Marketing, and Tech. Once sourced, the researchers conducted two forms of analysis on the corpus to determine the prevalence of gendered language.
The first analysis used Linguistic Inquiry and Word Count software (LIWC; Pennebaker, Boyd, Jordan, & Blackburn, 2015) to code the job advertisements for the presence of gendered wording using the word lists previously developed by the researchers (Gaucher, Friesen, & Kay, 2011). These words were also drawn from social science literature on gender stereotyping and were not specific to job advertisements.
The second analysis conducted by the researchers used a newly created program (DG Stat – created by senior software developer Joren Gaucher) to generate the most common and unique words in the specific corpus of job advertisements. All statistical tests were conducted using IBM SPSS software.
Before the data collection process began, the researchers also conducted a power analysis to determine an optimal number of adverts to collect per sector to determine the smallest sample size required to detect an effect within the given sample. The researcher’s previous analysis had found an effect size of d = .26 (Gaucher, Friesen, & Kay, 2011). This meant that for 95% power to detect a difference in gendered language wording between sectors, a minimum of 400 job advertisements would need to be analysed per sector. The researchers therefore conducted analysis on 500 job advertisements for each of the ten sectors listed above, equalling to 5,000 job advertisements in total.
Established in the UK in 1976, PageGroup has grown to become one of the world’s best-known and most respected recruitment consultancies. With operations in 36 countries, PageGroup provides recruitment services and career opportunities on a local, regional and global level.
PageGroup operates four key brands: Page Executive, Michael Page, Page Personnel and Page Outsourcing, supported by additional brands across the business.