Why is it that women are consistently paid less than men?

Worstall @ the Weekend There is undoubtedly a gender pay gap in our society. Even a number of them across our various societies.

In the US the usual number given is that women make 77 cents on the dollar that men do and here in the UK we've recently had a government shouting that the part time gender pay gap is as much as 37 per cent.

This could indeed be a societal scandal or perhaps we should come over all Kinky Friedman (a master troll, shurely -Weekend Ed) and start shouting anti-feminist remarks by telling women to get their biscuits in the oven and their buns in the bed. And stop worrying their pretty little heads over the natural order of things.

Or, possibly, and I know this would be a rarity for opinion journalism, we could attempt to work out whether there is a pay gap, if there is what's causing it and thus whether we think it's a reasonable outcome or not. This is all prompted by a request from Banksy here.

The opening point has to be that the 77 cents number is correct (the 37 per cent is statistical manipulation from Labour's acting leader Harriet Harman), so our next task is to work out whether this is a reasonable or fair outcome or not. To cast it into the usual economic terms, we want to know whether we're looking at taste discrimination or rational discrimination.

Are women being paid less just because men are pigs or because there is some rational reason why women might be making less money than men?

As American economist Gary Becker pointed out, we don't think there's all that much room in a market economy for taste discrimination: if you're refusing to hire good workers, or deliberately underpaying them, then there's room for someone to come in and hire them cheaply and outcompete you.

And there's no doubt at all that decades back this was true, as Dame Stephanie "Stevie" Shirley was able to prove with FI Group. Hire those female programmers with children that no one else will and make a fortune. Our question today is different: is that room still there?

The American numbers have been well crunched by Christina Hoff Summers and you can follow that here. I've been following this for years on the UK numbers so we'll investigate those. Partly on the grounds that I know them better and partly on the grounds that this case, built up over those years, has actually led to a change in UK law.

While the specific numbers differ slightly, the underlying story is much the same.

The one story that is consistent with the maximum amount of evidence is that there isn't actually a gender pay gap of any size in our modern society. There is a significant motherhood pay gap. To the extent that you want to equate the ability to be a mother with gender you can still call this a gender pay gap, if you wish. Although that seems to run counter to other modern mores.

Our starting point for numbers needs to be ASHE. Mean and median hourly, weekly etc wages, broken down by race, gender, age cohort, full and part time and so on.

We do have to be extremely careful about these raw numbers, though, as they don't tell us quite what we might think.

It's long been an observation that the highest hourly wages in the country go to black women. No, really, they do. Not because this is a society that well rewards melanin content plus ovaries in the workplace, but because the vast majority of the black women who work in the UK do so in London, where wages are rather higher than the rest of the country.

Crunching through these numbers we find that there is no gender pay gap among people under the age of 30. There is a small one among the next age group, and that persists, even widening out as we get up to people in their 60s.

That last figure is consistent with there having been taste discrimination in the past, as Dame Stevie showed there was. And I don't think that anyone's really going to argue that there wasn't for people who were educated or came of age in the 1960s and '70s.

But that there's no pay gap among the young does seem to show that we've not got that going on now.

Which leaves us with that age at which the gap does blow out. And given that the average age of women who become pregnant for the first time now stands at 30 or so, that the age at which the gap does expand has climbed, in both the UK and US, as that age of primagravidae has risen, would lead us to think that it is about children and the having of them.

We've also other bits and pieces that point to the same conclusion. Never married single women in their 40s actually have a pay gap in their favour, over men. Lesbians, who are less likely to be mothers, have higher pay than other women and – by some estimates – higher than men of the same age cohort. And when we look at very detailed numbers we find that there's not usually a pay gap for a specific job.

Here we cannot look at “women in finance” or “women in tech companies” because those are much too broad a category. To be a sexist pig about it, HR and front-line production are going to be paid differently in both industries, while we would all rather assume that there's going to be some gender segregation over which gender is employed in which sector of those industries. When we look at job for job properly, we really don't see gender pay gaps.

There's a couple of times we do, nursing seems to pay women better than men (consistent with Simon Baron-Cohen's ideas on systemising and empathising) and jobs where sheer physical grunt suits the male physique tend to pay men better. But other than those there's not much in it in any really direct comparison.

So, we end up with really thinking that it is children and the having of them that produces that general difference in pay. Something that again we can check: in ASHE, we can see that full-time female employees work fewer hours on average than their male counterparts. And there's a very much larger number of women working part-time than there are men.

This is consistent with the idea that, having had children, women find that raising them is of possibly more interest than just slogging it out at work. Perhaps that in itself is disgustingly sexist, although I can't quite bring myself to think of it as off in a viviparous mammalian species.

Pregnant pause

We can also test this in another manner, by looking at cross country comparisons for the gender pay gap. Here are the EU figures, and it's notable that the raw gap (i.e., that 77 cents idea) is pretty much the same for the UK and such hotbeds of gender equality as Sweden, Norway and Iceland.

One extraordinary outlier is Italy, with a gap near one third of those countries and it's really difficult to think of that as a country outstanding in its gender equity compared to the Nordics. The explanation is that it's most unusual for a married Italian woman to work, especially once she's had kids.

Thus, rather than going back to work at lower wages to offset the desires of schedules and time off and so on, they just never do re-enter the workforce. Those that do tend to be the upper middle classes working in careers, thus dragging up the average wages.

We can also now explain Harman's claim of a 37 per cent part-time gender pay gap in the UK. The actual part-time pay gap between men and women is in favour of women. If we measure women who work part-time and men who work part-time (everyone here is working from those ASHE numbers).

But all part-time workers make less than all full-time workers ... well, not quite all, but the averages certainly work out that way. So, what Harman did was calculate the wages of part-time women with respect to full-time men.

That's a calculation that led to Sir Michael Scholar, the statistics board bod, rapping her over the knuckles for being misleading.

It is of course possible to be mischievous on this subject, as I once was for The Guardian. This was back in the days when that newspaper would occasionally bring in someone to poke the readers with a pointy stick.

I noted that, yes, there was this motherhood pay gap. But don't worry, because society already offers child tax credits and child benefit, the two amounting to about what that motherhood pay gap was. So, mothers are already being compensated. But agreed, that really is pointy stick through the bars stuff.

I would also point out that I've been offering this rough analysis around for some years now, and I know which political activist I convinced that at least part of the problem lies in maternity leave and maternity pay. Mothers do take time off to care for the little darlin's, this does interrupt careers and this will be an explanation for some of the subsequent pay gap.

The political activist in question persuaded Lib Dem and former deputy PM Nick Clegg of this and that's why the country now has paternity leave, so that it becomes at least somewhat a parents' pay gap, not a motherhood one.

Perhaps not quite the policy choice I would have made (“Suck it up, honey!” being more my attitude, you want the blessing of children then there's costs that go with them) but it is at least a policy that starts from the reason for the numbers being as they are and then attempts to change the outcome sensibly.

As best we can tell from the numbers available to us there's somewhere between a very small and no “gender pay gap” in the sense of women being paid less than men simply because they're women. Or, in the economists' jargon, no to very little taste discrimination.

There are different outcomes in average pay as a result of life choices that are being made by people. Some of them are pure biology, others are societal in the expectations of who is the primary care giver. We can call that second “gender discrimination” if we choose but some very large part of that to damn near all of it is that mothers do seem to prefer to be that primary care giver themselves. House husbands do exist, the choice is open, and it's a rare enough choice.

Finally, a provocative little thought. Not a pointy stick and bars thing, rather an observation.

On the other side of these numbers there's a well known phenomenon of fathers making more than non-fathers. In part this is an age cohort thing, the older you are the more likely to you are to be a father. But it persists even accounting for that.

And that fathers to non-fathers pay gap is rather larger than any residual we can find in the male/female pay gap after we've accounted for life choices. My own supposition here is, and it's entirely a supposition, that this is the last little building block that we need to make sense of the whole outcome. Men and women, on average, react differently to the having of children. Men seem to, at the margin, do more work to provide for them, women a little less work at the margin to take care of them.

And ain't that just such an exotically surprising finding in a mammal species? ®