Over 20 years ago, I began my first job search and quickly discovered that landing a job in Colombia was not an Amazon Prime experience. I submitted countless resumes with my photo to job advertisements that solicited women only and after getting the job offer, I was sent to a local hospital to get a pregnancy test. I received clearance and was offered the job. Nowadays there, most people go through psychological testing, background checks, and home visitations to get hired for professional jobs.

At the time, I did not realize that the hiring systems were plagued with gender & racial biases. Why am I sharing this? Because as you can see in the image below, we are in 2020, and organizations continue to use the same biased processes while trying to recruit a diverse workforce. Last year, a tech company was found guilty of asking for “preferably Caucasian candidates” in its LinkedIn job advertisement. After reading this, I thought if we cannot rely on organizations to remove biases from their job postings, could LinkedIn become a leader in making it happen?

The short answer is YES, if LinkedIn invests in some product redesign. Let me explain: when I got to the U.S., I was surprised & relieved to not have to submit a picture with my resume. However, all of that has changed with the entrance of online marketplaces. A field study of Airbnb’s online market place found that the lack of anonymity led to a 16% rate of discrimination against blacks. I thought if the same were to be true on the LinkedIn platform then what is the negative impact on the global labor market? I analyzed LinkedIn to uncover areas for improvement with the hope that they will examine those areas & positively shift global hiring practices to create a more equitable labor market.

What design features could be added or changed to eliminate bias?

  1. Job Description Language

Most job seekers begin their journey reading job descriptions and unconsciously carrying biases that influence their decision to apply. I analyzed language in three hundred CEO & software development job descriptions from the tech sector using Applied Technology and found a large percentage of senior positions are written using a masculine tone, despite a general trend in the market toward neutralizing gendered language. These results are consistent with a longitudinal study conducted to analyze gender bias on seventeen million job postings. Although the study found that changing the masculine words (E.g. assertive, ambitious, competitive) to gender-neutral words (Eg. gentle, warm, understanding) led to a small effect on women’s decision to apply for the jobs, we should not allow job descriptions to perpetuate gender stereotypes. LinkedIn could add a text analysis feature to help employers reduce the amount of gendered language embedded in their job descriptions, and attract a more gender-balanced group of candidates.

2. LinkedIn Recruiter/Talent Solutions

Based on my experience, I know that it’s a common practice among recruiters to visits a person’s LinkedIn profile to quickly assess fit for the position. I began to question this practice and its impact on candidates after reading research showing that resumes with black names receive 50% fewer callbacks than whites in the U.S. labor market. The issue is that LinkedIn algorithms reward Profiles with photoswith more views and messages versus generic ones. For candidates subject to gender, age, or racial bias, this ubiquitous practice among recruiters can add one more layer of discrimination.

LinkedIn could reduce the impact of unconscious bias by providing a list of qualified candidates to recruiters without including the profile photo. Take it one step further and anonymize names until the recruiter screens profiles for qualifications alone.

3. Find the Right Candidates Faster

This LinkedIn Recruiter feature “Find more people like” lets you create a search based on ideal candidates you may know. In theory, this seems like a fantastic way to save time and quickly find new candidates, but in practice, it can lead to a highly biased candidate pool. In my work analyzing hiring practices, I find that most hiring managers tend to frequently cut and paste old job descriptions or write new ones based on the qualifications and tasks executed by the last person performing the job. Then, recruiters use the job descriptions to search on LinkedIn for possible candidates. It’s like looking for a new partner using your ex’s tinder profile, it may lead to an interesting pool but not a diverse pool of candidates. Thus, it will be critical that the LinkedIn algorithm ensured gender and racial representation in its suggestion of candidates matching the selected criteria.

4. Number of Candidates Who Have Applied

According to this study, women prefer to apply to jobs where they see low competition as they tend to focus more on cooperation and may question their fit for roles more often. As you can see below the LinkedIn job board provides candidates with data indicating the number of applicants to a given job. This data point can potentially skew the applicant pool more male than female. Removing this data point could make a difference for companies seeking to attract more female talent in a male-dominated industry.

5. Job Advertisement

A new study showed that Facebook’s job advertisements contained biased algorithms in which women were more likely to be shown supermarket clerk jobs while black men were shown taxi driver opportunities. Companies pay extra to advertise their jobs on LinkedIn hoping to attract a large pool of candidates. I find it interesting that my feed lacks C-Suite job level offers leading me to ponder the type of bias applied to my profile.

Let’s make the platform more transparent and provide data that shows the demographics of the population seeing the ad to avoid segmenting the candidate pool & contributing to bias like what’s been documented on Facebook.

Finally, my dream will be to never see again a job advertisement asking for a male or female candidate. I’d like to see LinkedIn emerge as a leader in eliminating bias and increasing representation in the global labor market. This is a call for their leadership to achieve their mission by helping LinkedIn professional users connect to an equitable labor market & achieve true success.

Note: This article was first published on Linkedin.


  • Edelman B. et al. Racial Discrimination in the Sharing Economy: Evidence from a Field Experiment. American Economic Journal: Applied Economics, Vol 9, N. 2, April 2017, (pp. 1–22.)
  • Shiliang Tang, Xinyi Zhang, Jenna Cryan, Miriam J. Metzger, Haitao Zheng, and Ben Y. Zhao. 2017. Gender Bias in the Job Market: A Longitudinal Analysis. Proc. ACM Hum.-Comput. Interact. 1, 2, Article 99 (November 2017)
  • Bertrand, Marianne and Sendhil Mullainathan. “Are Emily And Greg More Employable Than Lakisha And Jamal? A Field Experiment On Labor Market Discrimination,” American Economic Review, 2004, v94(4, Sep), 991–1013.
  • University of Michigan. “Women’s preference for smaller competition may account for inequality.” ScienceDaily. ScienceDaily, 12 May 2016. <www.sciencedaily.com/releases/2016/05/160512142920.htm>.
  • Facebook’s Ad System Might Be Hard-Coded for Discrimination. https://www.wired.com/story/facebooks-ad-system-discrimination/