Is the gender pay gap caused by discrimination, or the different choices men and women make in the labour market? Unless you are at either extreme end of this debate, you are unlikely to think it is wholly one thing or the other. But the truth is that we are in the dark about where the balance lies. No wonder that policymakers across the world are reaching for their flashlights.
In the past five years, Australia, Japan, Germany, Lithuania, Sweden, Switzerland and the UK have all proposed or implemented policies that compel companies to analyse and reveal their gender pay gaps. Will this sudden push for transparency help answer the question?
Two challenges are beginning to emerge. The first is to extract enough useful data from companies to shed light on anything at all. The second is that — as revealed by a study into the gender pay gap among Uber drivers — we need to ask a smarter question.
The UK typifies the first problem. The government has told companies with more than 250 employees to publish their mean and median gender pay gaps by next month. Some have warmed to the idea more than others. More than 30 companies claim to have both a zero median and zero mean pay gap, which is statistically implausible. One company says its median gap is 116 per cent, which is impossible. Five out of every six companies have not yet complied.
Patchy compliance in the UK is a shame. Across the Atlantic, Uber has demonstrated how valuable a fine-grained analysis can be. The ride-hailing company co-operated with academics to investigate the pay of more than 1m drivers in the US. On the face of it, it is not obvious why there should be any gender pay gap at all.
The algorithm that assigns customers to drivers does not know the drivers’ gender. The fee-per-trip is the same for everyone. There are no constraints on working hours. Yet the study found an earnings gap of 7 per cent more per hour on average.
The researchers found three factors which fully explain the gap. Firstly, men drive about 2.2 per cent faster than women, which allows them to complete more trips per hour. Men also have more experience on average: they drive more hours per week and they are more likely to have been with Uber for longer. Drivers who have completed more than 2,500 trips make nearly 14 per cent more per hour than drivers in their first 100 because they have mastered tactics to maximise earnings. Finally, men tend to drive in more lucrative parts of town.
The researchers’ conclusion? Even in the “gender-blind” gig economy, women earn less because of their “gender-based preferences” about where, how much and how fast to drive. In other words, women earn less because of the choices they make. But the language of “choice” and “preference” is tricky. Some choices are made more freely than others.
The speed at which you drive seems to me like an unconstrained choice (although Uber might usefully do some more research into whether its pay structure is rewarding the right behaviour, in terms of passenger safety.) But I wonder if women choose to avoid some areas, even though they are lucrative, because they feel unsafe there. If that is so, is a choice between earning less or being unsafe really the same sort of choice?
Last year, a female Uber driver in the UK filed a sex discrimination claim saying the company had not done enough to protect women at the wheel. When I spoke to her, she told me that she did not feel safe driving at night and so missed out on surge pricing. She felt she couldn’t ask passengers to get out when they seemed threatening, since she thought it could damage her customer rating. Her lawyers claim this amounts to indirect discrimination: when the same rules applied to everyone disproportionately disadvantage one group (in this case, women).
To be clear, this accusation is not proven. The UK case has not yet been heard and there was not enough detail in Uber’s US study to explore the possibility that female safety was a factor. Uber says female drivers feel much safer on its app, where they are tracked, than in a traditional taxi. But the wider point is that we should not assume that all “gender-based preferences” are something employers cannot (or should not) do anything about.
Sometimes women’s choices are freely made. But sometimes they are shaped by societal constraints that an employer might be able to address. Uber cannot help the fact that women are more vulnerable to sexual violence than men, but it might be able to build extra safeguards into its app to make its female drivers safer. A law firm cannot help the fact that many women still do more unpaid care work than men, but it can design senior jobs so that they do not automatically rule out people who cannot work 50 hours a week.
Any employer grappling with the gender pay gap should think carefully about the choices female staff make about their work, and whether they make them happily, or because of constraints the company could do something about. If in doubt, there is an easy way to find out. Just ask them.
Copyright The Financial Times Limited . All rights reserved. Please don't copy articles from FT.com and redistribute by email or post to the web.