An end to excuses for paying women less

Posted on September 13, 2016

Gender pay gap...File photo dated 11/8/2014 of office workers in central London. Less than one in 10 firms publish gender pay gap information, so face a "significant challenge" in meeting planned new reporting requirements, a study has shown. PRESS ASSOCIATION Photo. Issue date: Wednesday September 2, 2015. Research by the EEF manufacturers organisation found that less than a third of companies were prepared for the move to publish data, currently being consulted on by the Government. See PA story INDUSTRY Pay. Photo credit should read: Philip Toscano/PA Wire©PA

Every debate about the gender pay gap seems to end up in treacle. The numbers are the problem. Most statistics on the average gap between male and female salaries are met with a volley of “yes buts”. Yes, but does that control for the fact that more women work in lower-paid sectors? Yes, but does that control for the fact that many women take time out to have children, so accrue less experience?

The answer is usually “no”. And so the stalemate endures. One camp says the gap is a “myth” because it is the result of women’s own choices. Another says the very fact there are fewer women in senior jobs, or fewer women in high-paying sectors, is evidence of unfairness. A third camp says some women are victims of direct discrimination, paid less than men to do the same jobs (which is illegal in many countries).

    A new study from the London School of Economics unsticks the debate, at least for one institution: itself. The university commissioned Oriana Bandiera, one of its economics professors, to investigate its own gender pay gap. She took anonymised payroll data for LSE academics, controlling for age and length of service. She also controlled for the quality of their research using “REF score” predictions — the LSE’s assessment of how well they would do in the external system that assesses the quality of research. She found women were earning 11 per cent less, on average, than men of similar age and tenure, producing research of similar quality.

    The difference was negligible among the lowest paid but 30 per cent among the highest paid. This is partly because women are under-represented in high-paying departments such as economics. But, after controlling for department, the average gender pay gap was still 6 per cent. It has also grown wider since the late 1990s (although those data lack the control for quality because REF scores are a recent phenomenon).

    Prof Bandiera cannot control for everything and REF scores are not a perfect measure of performance. But at least these data allow you to zero in on the interesting questions. She suspects one reason for the widening gap is that academic salaries have grown more responsive to offers from rival institutions, particularly at the top. She speculates that men can more credibly threaten to move because they are more likely to have partners willing to follow them. “It’s a lot more credible for a man to say: ‘Stanford made me an offer, I’m going,’” she argues.

    This would suggest that relying on the open market to determine pay can create inequities that reflect not just the value of the employee but also their ability to leverage it. That might affect all sorts of people, not just women. The LSE has now asked heads of department to nominate academics for pay rises, with a particular focus on women.

    From 2018, the UK government will compel big companies to follow suit and publish their gender pay gaps. But they will be required only to publish median and mean gaps, together with the number of men and women in each pay quartile. This is too little information. It will lead to another flurry of “yes but” questions without satisfying answers.

    Far better try to control for as much as possible. Every employer ought to have data on age, tenure and performance. They might find very different patterns from the LSE. The Bank of England
    said on Wednesday that its mean and median pay gaps were 19 and 26 per cent respectively. Helpfully, it also published the gaps within each salary scale, which were much smaller, though not non-existent, at higher grades.

    Only when you understand the various sources of your pay gap can you decide how (or indeed whether) to address it. The message for any employer that says it wants to tackle the gender pay gap is simple: put your spreadsheet where your mouth is.

    sarah.oconnor@ft.com

    You must be logged in to post a comment Login