Wednesday, 27 January 2010

Do men really earn more than women?

Everyone's favorite minister has today announced another Government sponsored report looking at Inequality (details here).

Shockingly, the report finds that women earn less than men. Luckily the Government is now forced to provide as much data as possible to back up their claims, so I was able to delve into that data and see if I could replicate some of their graphs.

As I'm not a complete idiot, I understand the necessity of only comparing like to like. So for this piece of analysis I'm only going to look at Full Time workers. There might be a gender disbalance for the ratio of part to full time workers, so I'm not going to bother correcting for that and rather than ignoring it, I'm excluding part-timers. Also (mainly because I'm lazy), I'm going to investigate Median earnings, as this is usually considered the best measure (it excludes the extremely high and the extremely low earners automatically).

Right. Headline figures:
Median Weekly Earnings for Men: £494
Median Weekly Earnings for Women: £386

Gender Pay Gap: Men earn 28% more than Women.


That's pretty clear , lets have a look at how that varies by age:

As we can see, even at 16-19 Women are earning less than Men. The gap widens all the way up to the 50s then slightly falls. At every age group the Median woman is earning less money every week than the Median man.

Then I had a look at how Hourly Wages compare:


Here the picture is subtly different. Whilst there's still a big pay gap between the ages of 35 - 59, at younger and older ages it's almost entirely vanished. In fact, the Median 25-29 year old woman has a higher hourly wage than the median man! It appears the massive pay gap we saw earlier has mostly vanished.

Of course, the obvious reason for that is that the hours worked by Men and Women are different (although both are working full time). So I worked out the inferred hours worked per week by our median man and woman (Are they a couple?).

Wow! At every single age group, men are working for longer hours than women. This increases as both get older. The median man (across all age groups) is working 4.5 hours a week longer than the median woman. A 65-69 year old median man is working over 5 hours a week longer than his female counterpart.

This is actually quite fun news, because it gives us something we can control for. We can increase the weekly earnings for women by assuming that they're all suddenly going to become 'hardworking' and work the same hours as the median men are doing. This gives us the following graph:

Our "Hardworking" woman are now much more in alignment with the male earnings. There is still a big gap, starting from around 30. Coincidentally, this is almost the average age a woman has her first baby.

What are the headline figures now?
Median Weekly Earnings for Men: £494 (the same as before)
Median Weekly Earnings for Women: £431 (higher due to more hours worked)
Gender Pay Gap: Men earn 14.5% more than Women.

The Gender Pay Gap has halved. Just by correcting for a single metric (hours worked) I've singlehandedly done more to enhance women's pay than the last 12 years of New Labour.

This still leaves us with about a 14.5% difference though. Can we explain why this would be? I'm certain that in the vast majority of industries Gender makes no difference to pay. I earn the same as my female colleagues at the same level and it's very hard for a company to do otherwise. Clearly there is another factor at play.

I actually mentioned what I believe it is above. Babies. The average woman has about two children, starting on average at 29.

I've built a very crude model. It assumes a man and a woman both start work on the same day and earn the same. Both get 5% pay rises each year until they retire. The man takes no Leave of Absences, so works flat out from 20 to 60, then retires (hopefully to a nice civil service pension). The woman does exactly the same, with one exception. She doesn't get a pay-rise for three years as she's taking maternity leave (knocking a minimum of six months out of each year) and takes time to get back up to speed in her old job.

Over his lifetime the man earns £1.3m. The woman earns: £1.1m.

The difference is 14%, almost exactly that actually found in the survey above (If you assume two years of no pay rises, then the difference is 10%).

Clearly most, if not all, of the current Gender Pay divide is actually down to women choosing to delay their careers and take materinity leave! If the government was serious about reducing the divide, then they should mandate equal paternity leave (as they do in some Scandinavian countries).

The one thing these figures show is that once you correct for women choosing to have babies and choosing to work shorter hourse, men and women earn almost exactly the same. Why then do we need further measures to ensure 'equality'?

2 comments:

  1. A client of mine recently discovered that, despite being the No.1 best-performing salesperson in her team (and the only woman), she earns about $200 a week (20%) less than men who do a worse job (several of whom have been relieved of the burden of their employment there).

    I'm currently advising her on how to remedy the situation.

    You make a good point about "once you correct for women choosing to have babies and choosing to work shorter hours, men and women earn almost exactly the same" - but what about women who do not have children (by choice or circumstance) - should they be reimbursed the difference?

    ReplyDelete
  2. My figures are UK only, where I believe the law is stronger than the US.

    I think there are very few situations in the UK where a woman earns less than a man for the same job (with the same performance). Those will only be in jobs where salaries are very negotiable and hidden.

    My point was that once you apply the correction for women who tend to work less, they actually appear to earn the same. For the women who don't have kids (and work the additional 4 hours more per week on average), they'll probably be earning exactly the same as their equivalent males. I see no need to 'compensate' women for their gender to correct for a pay-gap because I don't beleive that those women will see a pay gap. Only those that had children and work fewer hours will see a gap (But of course there are obvious exceptions).

    I don't think your client should be reimbursed. I think your client should be paid a fair wage for her superlative work - if she then chooses to work fewer hours or leave employment then her wage should be expected to fall.

    Just in case it's not clear - Any situation where a woman earns less than a man for the same work and performance is reprehensible.

    AW

    ReplyDelete