Using AI to help eliminate the gender pay gap

Women must work 3 months more to earn the same as men

April 2 was Women’s Equal Pay Day, a day which symbolizes the additional months in the following year the average woman must work in order to earn what her male counterpart earned in the preceding year. But what does this mean, exactly? Look around and you’ll see many examples of women and men in the same positions getting paid the same wage. In many places, it’s illegal to discriminate based on gender. But this doesn’t necessarily translate to women and men being treated equally when it comes to their careers. Now, companies are starting to turn to AI to help reduce the gender gap.

Gender parity in the workplace: A murky issue

Start an argument on gender equality and you’ll soon find yourself mired in foggy arguments and emotional anecdotes about companies that hire inept women over hard-working men, just to fill their quotas. And in many ways this is true – the steps most companies take to fulfil their diversity goals aren’t always the fairest.

While we all know that there’s a wage gap between women and men, it’s hard to understand exactly how it works. Justifiably, men get angry when they see job postings for “female-only” candidates. How do you increase female participation in your hiring practices without turning the tables and discriminating against men?

Still other people refuse to acknowledge the problem. A poll by SurveyMonkey found that half of male millennials think women and men are paid equally – indeed, many men and women in the same roles are paid equally. For those who acknowledge the gender pay gap, a majority believes it will disappear in their lifetime.

The glass ceiling is real, and it’s not getting any better

While many companies have taken vows to aim for gender parity, flaunting their new female CEOs and egalitarian hiring practices, the majority are finding it harder than they imagined to make a real, lasting difference.

A recent report released by the World Economic Forum found that if we continue at this rate, female workers won’t receive equal wages until 202 years from now. Pew Research backed up this claim, saying the gender pay gap has more or less remained the same since the 1980’s. Despite the marches, the evidence, the corporate and government initiatives, the #metoo movement, the feminist literature and cinema, companies have not figured out how to pay women equally.

Why is it so hard to pay women equally?

New findings indicate the problem might be largely due to crossed wires between the HR and payroll departments. While female workers and male workers in the same positions often earn the same, some critics argue that female applicants with the same amount of experience as their male counterparts are often hired into lower-level roles.

This effectively results in a salary cut, even if on paper it looks like the female workers are earning the same as their male counterparts. These discrepancies get amplified throughout the promotion life cycle, resulting in the famous glass ceiling and the lack of females in higher positions.

How can AI help bridge the gender pay gap?

Interviewers harbor unconscious bias; often, biased decisions are purely caused by human weakness. Many AI hiring algorithms are also biased. Algorithms spit out analysis based on the data we feed them, and the data so far says that historically, the ideal candidate for most well-paying jobs have been male.

The lack of transparency in the hiring process is further complicated by the fact that companies tend to keep their HR and payroll data on separate systems. Information about a woman candidate’s MBA degree and fifteen years of experience as CEO at a flourishing firm gets mired in the HR system, and the payroll algorithm continues on its merry way, assigning the woman a salary and benefits based on what an average person in that job position should be earning.

AI provides data-driven suggestions on reducing the gender gap

Gapsquare is an AI company that aims to fix this problem. The company has found that after accounting for performance, years of service and education, there remains a 60% unexplainable gap between men and women which is the direct result of bias and societal norms.

To help beat this inequality, Gapsquare combines payroll and HR data into one system, enabling comprehensive comparison across different levels. AI and machine learning technology then provide an analysis of pay discrepancies based on demographic factors such as gender and ethnicity, along with data- and academic-based insights on how the discrepancies might have originated and a list of possible solutions.

Another AI-based company looking to eliminate the gender pay gap is the aptly named Pipeline. Based on research conducted on 4000 companies in 29 countries, Pipeline found that men were promoted 30% more often than women, and that the glass ceiling started lower down than one would expect. In an interview with Fast Company, Pipeline’s cofounder Katica Roy said the while it’s important to give equal pay for equal work, equal opportunity for career advancement is even more important.

Pipeline monitors the internal workforce data on an ongoing basis and provides an analysis every time a performance review, promotion or salary decision crops up, along with advice for closing the equity gap. The software uses natural language processing to provide non-biased reviews.

Gender parity makes economic sense

AI provides the transparency and objectivity that is often lacking during the hiring process and in the deliberation of pay structures. The algorithms can sift through factors like pay history, employee performance history, education, experience and other trends while ignoring variables such as race or gender, in turn reducing the unconscious bias that is often present.

HR-focused companies like Gapsquare and Pipeline that use AI to reduce the gender gap are gaining ground in industries from hospitality to tech and financial services. Slowly, slowly, we are starting to see more women and minorities in higher-up roles, and there is strong evidence that gender parity in the workplace boosts economic growth. Pipeline, for example, has found that as gender parity increases by 10%, revenue increases by 1-2%. Companies can expect returns in about three years if they follow the AI’s recommendations.

AI is often touted as a worrying example of an industry with a staggering gender gap. Ironically, it might be AI that helps to one day eliminate the gender gap.