ways to counter recruitment bias against women

ways to counter recruitment bias against women

Recently, Amazon’s machine learning experts uncovered a troubling problem: their new recruiting engine was biased against women. The engine had taught itself to prefer men - based on its evaluations of resumes received from a predominantly large number of male candidates. While this underscores the limitations of using technology for recruitment, humans are not faring well either. Conscious and unconscious human biases are holding women back at work.

A Yale University study found that both male and female scientists (who are trained to be objective) were more likely to consider men to be more competent than women, and pay them $4,000 more per year than women. Bias doesn’t end with recruitment. It follows women throughout their careers. In 2018, the number of female chief executives fell by 25% in Fortune 500 firms. The staggering fall is disheartening as the number of women in top roles is already dismal. Women are 13% less likely to be promoted to the manager level and 26% less likely to move to a senior vice president level. The disparity reaches its highest at the boardroom level - with 31% fewer women making it.

How gender bias creeps into the recruitment process

While studies have shown that a gender diverse workforce is good for businesses, it is difficult to completely avoid unconscious biases - such as favoring people who look like us or biases that result from our desire to conform to the majority. Harvard economist Iris Bohnet presents an interesting concept in her book What Works: Gender Equality by Design . According to her, "Seeing is believing". The fact that we don't often see male kindergarten teachers or female engineers means we don't naturally associate women and men with those jobs, and unconsciously apply different standards when hiring, promoting, and evaluating job performance.

Similarly, cultural biases can lead people to assume that men make better leaders - as men are thought to have leadership qualities like gravitas, while women are believed to be better at supporting roles because of their dependability. In a recent experiment, when asked to ‘picture a leader’ - both men and women drew images of men. Industries that pay more, such as manufacturing, oil and gas, engineering, and more recently, roles such as data science and analytics, also favor men when it comes to talent acquisition. Such biases ultimately increase the gender pay gap.

A recent study found two key types of discrimination that tilt the career game in favor of men:

  • Taste-based discrimination This form of discrimination is driven by stereotypical thinking and favoritism towards one sex.
  • Statistical discrimination For risk-aversion purposes, a decision-maker may substitute an information void related to an individual’s productivity, with group averages (real or imagined) or stereotypes. While taste-based discrimination is visible and perhaps easier to weed out, statistical discrimination isn’t obvious even to those practicing it, making it far more difficult to side-step.

How HR practitioners can counter gender bias in talent acquisition

Here are a few strategies that recruitment leaders can leverage to de-bias their practices and procedures:

  1. Re-engineer your job descriptions Gender discrimination can start even before a candidate is interviewed or hired in the form of biased job descriptions (JDs). A recent study revealed that stereotyping is prevalent across sectors in varying degrees. JDs for social care are 87% more likely to use female-biased language, as are 67% of JDs for admin-roles. In contrast, sales and management roles have a 16% and 10% higher likelihood (respectively), of using male-biased language.  Australian software company Atlassian increased its female technical hires from 10% to 57% (in just two years) by leveraging an online writing tool to weed out gender-biased language. Gender-neutral job descriptions also reduce time to hire - by increasing the talent pool – as they result in 42% more applicants.
  2. Adopt blind resume reviews The Boston Symphony Orchestra in the 1970s resorted to blind auditions to increase the representation of women and found that it increased the chances of women being hired by 25–46%.   Adopt a similar approach by leveraging tools to remove all gender markers from a resume such as photographs, names or biographical information. The result: ability to hire based on far more critical and relevant criteria such as performance and cultural-fit.
  3. Test job performance beforehand Work sample tests that determine how a candidate would solve problems or react to situations that occur commonly on the job is a great way to evaluate candidate potential. To retain the focus on performance and capabilities and reduce the chances of hiring bias, leverage scientifically designed work sample tests that evaluate applicants’ domain knowledge while eliminating bias. Leading companies like McKinsey use work sample tests based on real client scenarios to de-bias hiring.
  4. Institutionalize gender diversity Push to create a formal organization-wide gender diversity policy – an area that only 66% of companies currently pay attention to and even fewer actively monitor (25%). Use data to help strengthen your case for instituting such a policy. Several studies show that a larger percentage of women in executive positions ensures 34% higher returns to shareholders, 26% higher return on invested capital, and superior competitive advantage.

In order to make gender diversity a part of the core values of a company, it’s important that employees also understand how gender bias works. However, only 67% of companies offer employees training aimed at eliminating hiring bias while only 56% train employees to weed out bias in performance reviews. To be successful, gender diversity training must be comprehensive and include insights on how gender bias creeps in, why eliminating it is essential for businesses and what the best strategies are for attaining this goal.

subscribe

Fields marked with an * are mandatory fields
  1. *
  2. Submit
  3.