The gap between statistics and discrimination

The gap in pay between men and women has narrowed, according to the latest data from the ONS, no matter how you measure it.  

The Annual Survey of Hours and Earnings report, published last Thursday, showed a small but encouraging decline in the pay gap. Yet, on Tuesday The Guardian's front page had declared that “Male doctors earn £15,000 a year more than women”, based on a report from the British Medical Association.
The ONS' report showed that in the past year the median pay gap for full-time employees has decreased from 12.6 per cent to 12.2 per cent, with the women’s hourly rate rising by 4.3 per cent to £11.39 and the men’s by 3.8 per cent to £12.97. The mean pay gap also narrowed, from 17.4 per cent to 16.4 per cent.  
The new figures met with generally positive comments. But the Equality and Human Rights Commission, whilst welcoming the improvement, complained that “if this rate of decline continues, it will be another 17 years before women and men will be earning equal pay.” That assumes, misguidely, that only when the figure get to zero will discrimination truly be eliminated.
Discrimination is not the only reason, nor even perhaps the most important, for the gender pay gap. Even if (and hopefully when) all discrimination is eliminated, a pay gap will still exist. The reason? Men and women generally have different interests, skills and motivations that dictate the line of work they go into.
Men seem to be generally more motivated by money than women, meaning that a higher proportion of them work in sectors that traditionally pay more. Women are more likely to follow a profession that they enjoy but might not pay as much; nevertheless, in surveys women are generally found to be happier than men.
The biggest difference in motivations – and perhaps, opportunities - is in the choice to work part-time. Only 11 per cent of employable men work part-time, compared to 41 per cent of women. Part-time work pays less, so the fact that more women do it means that the pay gap jumps to 22 per cent when both part-time and full-time work is counted. (This is despite the fact that women are paid more per hour working part-time than are men – 2 per cent more, according to the 2009 figures, down from 3.7 per cent in 2008.)
How much of the gap is the result of discrimination and how much a matter of choice and opportunity is hard to measure. The sweeping judgements comparing all men and women in employment do not highlight discrimination so much as show that, unsurprisingly, there are general differences between the sexes. Thus, the different motivations between the majority of men and women could explain most of the current 12.2 per cent difference between median pay for full time work. We simply don’t know.
This is where the BMA report  is valuable because it compares the difference in pay between men and women doing the same job. It shows that the mean gender pay gap for doctors is 18 per cent (£15,245 per year), which is very similar to the ONS' figure of 16.4 per cent for all full-time employment, and provided The Guardian’s headline.  
But the report, by Sara Connolly of the University of East Anglia and Anita Holdcroft of Imperial College,, goes much further than this. Differences in pay rates in medicine could be caused by many factors. For example, women entered the profession later than men, so may be under-represented in more senior positions. Women may choose areas of medicine where pay is generally lower, or may choose to work fewer hours. Only by using regression analysis to correct for all these factors is it possible to separate differences in pay that can be explained, and those that cannot, so are likely to be the result of discrimination.
This is what the BMA report does. Among other interesting conclusions, it finds is that taking a career break – to have children, for example – does not significantly affect the salary gap. The authors conclude that the pay gap that cannot be explained except by “difference in treatment for the same characteristics” (discrimination by any other name) is 5.6 per cent for consultants (£5,500 per year) and 4.1 per cent for trainees (£2,000 per year). If this is correct, between one quarter and one third of the gender pay gap for doctors is caused by discrimination.
These findings ought to be a greater cause for concern than the gross figures produced by ONS. If the medical profession is representative of others, it would suggest that gender discrimination in the workplace is costing women approximately 5 per cent of their rightful salary.