Researchers are turning to data analysis to find out why women are still paid less and make less progress in their careers than men.

Despite progressive moves in some companies and cultures to level the playing field, the World Economic Forum’s Gender Gap Report 2017 found the divides between men and women – across health, politics, education and economics – had increased.

Image: Global Gender Gap Index 2017

On current trends, it would take a century to close the overall global gender gap across the 106 countries covered since the inception of the report, while the "economic gender gap" was unlikely to be closed for 217 years.

But in an era when technology promises the answers all our problems, could data help to uncover the true causes for the gender gap – and even close it?

Sensors and discrimination

In an attempt to understand the continuing gender differences in career outcomes, one set of researchers decided to bug a multinational business. They placed sensors on 100 employees for four months to detect if there were any differences in how men and women worked and were treated.

The small ID-like sensors relayed information about who talked to whom and the volume and tone of voice in conversations, though they didn't convey the actual words.

Women in the company made up roughly 35-40% of entry-level staff, but were underrepresented at each subsequent seniority layer and held just 20% of positions at the two highest levels.

The same difference

“We went in with a few hypotheses,” the researchers wrote. “Perhaps women had fewer mentors, less face-time with managers, or weren’t as proactive as men in talking to senior leadership.”

However, as the data came in, they said: “We found almost no perceptible differences in the behaviour of men and women. Women had the same number of contacts as men, they spent as much time with senior leadership, and they allocated their time similarly to men in the same role.”

Women and men also had indistinguishable work patterns, while in performance evaluations both sexes received statistically identical scores at each level of seniority.

Another frequently argued point, that women lack access to informal networks because they do not “spend time with the ‘boys’ club’”, did not stand close scrutiny. “We found that ... women were just as central as men in the workplace’s social network.”

Expectations big and small

Instead, researchers say their analysis suggests promotions were dependent on how women are treated. “This indicates that arguments about changing women’s behaviour – to ‘lean-in’ for example – might miss the bigger picture: gender inequality is due to bias, not differences in behaviour.”

They add: “Our data implies that gender differences may lie not in how women act but in how people perceive their actions. For example, consider female mentorship programmes that try to connect high-potential women with management. If women talk to leadership at similar rates as men, then the problem isn’t lack of access but how those conversations are viewed.”

Image: PwC, Female Millennial Survey (2015)

Expectations of behaviour outside work were as important. Women tended to leave the company in the middle of their seniority, after four to 10 years, perhaps to take parenting career breaks, though the data did not allow researchers to determine whether this was the case. “But,” they say, “we don’t think this changes the argument for reducing bias.”

Suggestions for addressing the amount of bias ranged from hiring from a more diverse workforce and looking at workloads so women are not forced to make a choice between “family or work”.

Data step in

Perhaps more importantly, the research suggests companies should move away from “anecdotal evidence or cursory surveys” to measure gender equality and instead use hard data, especially when trying to deal with specific problems, such as whether a particular corporate culture is to blame for suppressing women’s ambitions.

Other studies have also examined whether measuring performance metrics can overcome gender bias. Yet, experts say, even when performance information is available, research still shows men tend to be favoured over women of equal ability.

The writers of one survey, examining whether professional investment decisions were based on the apparent sex of fund managers, said: “We found that [an investment] recommendation submitted by someone with a very female name, like ‘Mary’, received approximately 25% fewer clicks overall than a recommendation submitted by someone with a very male name, such as ‘Matthew’.”

Corporate and global leaders are aware of the problem. As Paul Polman, Unilever Chief Executive, has said, research suggests that “some of the strongest forces behind persistent gender gaps are harmful social norms and stereotypes that limit expectations of what women can or should do. These outdated norms that discriminate against women are all around us – and they are deeply ingrained.”

But to overcome them, we first have to become more aware of them: and in the 21st century, data analysis may well have a crucial part to play in closing the widening gender gap.

To have your say on this topic, submit a question to be answered at Davos by a global leader.