Feminists cannot breathe easy over the pay gap
Tim Worstall has stopped merely criticising how the EOC calculates the part-time gender pay gap, and started denying that any gender pay gap exists whatsoever. I wish he would read the excellent ongoing series of posts on the wage gap over at Alas, a Blog by Ampersand (the comments are good too), which would show him that many of his doubts, questions, and objections are extremely familiar to feminists.
I am certainly not going to try to tackle all Worstall’s queries in one post, but I will attempt to refute his recent statistical argument for denial, which (as I understand it) runs as follows. If discrimination were disappearing, we would expect younger women to suffer less of a pay gap then older women. According to the available figures, the gender gap does not appear until after age 30. Therefore we can conclude that the ‘gender pay gap is over, solved’.
Even if it were true, there would still be the matter of society’s debt to all women over 30 who are suffering ongoing injustice. But in fact the naivety of this argument is breathtaking. Look at the pay gaps implied by the very pay figures Worstall uses in the relevant post:
| Age group | Men (£) | Women (£) | Pay gap (%) | Years women turned 18* |
|---|---|---|---|---|
|
* This is intended as an extremely crude proxy for workforce entry, and assumes the 50+ group ends at 60. Source: Office of National Statistics (ONS) Annual Survey of Hours and Earnings (ASHE). |
||||
| 18–21 | 6.20 | 6.07 | 2.10 | 2002–2005 |
| 22–29 | 9.35 | 9.29 | 0.64 | 1994–2001 |
| 30–39 | 12.42 | 11.36 | 8.53 | 1984–1993 |
| 40–49 | 13.24 | 10.50 | 20.69 | 1974–1983 |
| 50+ | 11.43 | 9.79 | 14.35 | 1963–1973 |
Laughably, Worstall summarises this as: ‘So for those under 30 there is in fact no (noticeable) gender gap in pay, for those in their 30s there is some and for those over 40 a considerable one.’ Perhaps Worstall is now earning so much that he wouldn’t notice if the government deducted another 2% of his current income in tax this year? Not all of us can afford to be so blasé.
Have a think about these figures. If we were to trust Worstall’s assumption that current age-group pay gaps represent mainly historical discrimination, then we would have to conclude not only that the gender gap in wages in 1963–73 was only two thirds that of 1974–83 (which is just conceivably possible), but also that the gender gap tripled between 1984–93 and 1994–2001 (which seems highly implausible). And since 16 to 17 year old girls earn 104.65% of what their male peers earn, we must surely conclude that there was a radical reversal in gender discrimination in the last year alone!
Another point about Worstall’s figures is that whereas the EOC uses mean pay, Worstall uses median pay. Now there’s nothing deliberately deceptive about that. The Government’s Women and Equality Unit also uses median pay. But in this case it does compromise Worstall’s conclusions. The median is excellent at telling you about the most ‘common’ value in a set of numbers because it is less responsive to unusually high values, like so:
| Number set | Median | Mean |
|---|---|---|
| {1, 2, 3, 4, 5} | 3 | 3 |
| {1, 2, 3, 4, 10} | 3 | 4 |
| {5, 5, 5, 5, 5} | 5 | 5 |
| {5, 5, 5, 5, 6} | 5 | 5.2 |
As the Women and Equality Unit explain: ‘Mean figures are often not favoured because they can be affected by changes to the earnings of small numbers of very high-earners.’ But this also means that median figures for earnings are relatively blind to elite (male) privilege, to ‘glass ceiling’ effects. When you reconstruct Table 1 with mean pay, things look rather different:
| Age group | Men (£) | Women (£) | Pay gap (%) | Years women turned 18 |
|---|---|---|---|---|
| 18–21 | 6.81 | 6.51 | 4.41 | 2002–2005 |
| 22–29 | 10.84 | 10.25 | 5.44 | 1994–2001 |
| 30–39 | 14.66 | 13.01 | 11.26 | 1984–1993 |
| 40–49 | 15.98 | 12.58 | 21.28 | 1974–1983 |
| 50+ | 14.43 | 11.86 | 17.81 | 1963–1973 |
Everyone’s wages are higher, but the gender pay gap is wider and the differences in the gap between age groups are less. Moreover, the implications of Worstall’s assumption about current rates of pay and historical discrimination are rather different this time because 18–21-year-olds are now doing better than 22–29-year-olds.
The better way to track historical alterations in the full-time pay gap is, of course, to compare current data with older data. When we chart such changes from 1998, when ASHE started, to 2005 (see Excel spreadsheet), we find that while the general trend for all age-groups is gradually downwards, the details of the trendlines once again vary significantly depending on whether we use median or mean figures:
Worstall’s assumption about age groups is peculiarly misguided in that feminists would expect a larger pay gap to be visible among older women since discrimination is cumulative, as Ampersand explains:
Discrimination in the workforce is usually is a matter of ‘cumulative causation.’ Among other things, this means that the effects of discrimination add up over a lifetime. So, for example, losing a single job offer or promotion usually won’t make a big difference; but dozens of such small losses over the course of women’s careers eventually add up to a big wage gap.
This is important, because it means we should expect the pay gap between men and women at the start of their careers to be small. The effects of discrimination build up gradually over time, and only becomes sizable once women have been in the job market long enough for the impacts of dozens of individual instances of discrimination to add up. So when … [researchers] look only at the pay gap among young workers, they’ve selected workers who have not yet been in the workforce long enough to have experienced the worse of the pay gap.
Worstall would probably object to this line of thinking, because he seems to believe that those who study the pay gap think the key thing is equal pay for the ‘same job’. But while women may sometimes be paid less for exactly the same job, the debate over equal pay is not, and has never been, limited to such discrimination. It has always been about the devaluation of women and their work within and without the home, about society is structured to restrict their opportunities for their free economic development. The year-on-year failure to properly compensate women for their work produces a cumulative degradation of their quality of life. This is why general figures for pay are relevant, this is why the occupational segregation of women into particular industries and part-time work generally is crucial, and this is why blaming it all on women’s ‘different choices’ is irresponsible. The fairer sex ‘make their own history, but they do not make it as they please; they do not make it under self-selected circumstances, but under circumstances existing already, given and transmitted from the past. The tradition of all dead generations weighs like an nightmare on the brains of the living.’
January 5th, 2006 at 4.07 pm
The point about the accumulation of discrimination is well made. Hadn’t thought of it that way.
The median/mean was a result of me forgetting which EOC had used in its figures. Mea Culpa.
You have, unfortunately, left out quite a large part of what I said. It’s linked through in the piece you have above. That there are two possible explanations of there being a gender gap at the older ages and not at the lower. One is that women who have children are discriminated against. The other is that we used to have discrimination and now do not.
We could also turn it round. What evidence would we be looking forif we though that we had managed to stamp out discrimination? Evidence of the lack of a gender gap in those younger people who have not suffered from it.
You’ll have to work a lot harder than the above to convince me that this isn’t the case. As the EOC themselves have pointed out, women are now a majority of those in higher education, the majority of newly qualifying doctors and solicitors. That’s certainly evidence of less discrimination even if not of all.
The second point, that women with children may or may not get discriminated against. Perhaps they are and perhaps it’s entirely reasonable. Are they less produtive than those who don’t have such an interruption in their working life?
January 5th, 2006 at 4.18 pm
Reading over the post at ampersand I don’t see anything there that invalidates my claims or numbers. Remember, I’m using UK numbers, over there US. She agrees that much of the wage gap in later years is because of the effect of children and absence from the labour force, for example.
One very useful test (but we’ll have to wait a few decades to see how it turns out) will be what happens when traditionally high paying professions like medicine and the law are predominantly female, will they still be high paying? If not, then that would be evidence that there is indeed discrimination against female majority workforces.
Doubt it will happen though.
January 5th, 2006 at 9.38 pm
Tim,
As ever, many thanks for promptly coming along to reply.
I agree what you said was sometimes more complicated than my summary here. But I’m afraid it was you who moved from raising doubts and discussing multiple explanations to the dismissive soundbite that there is no gender pay gap: ‘the battle is won’, and so on.
As to the explanation based on ‘women who have children’, I’m not convinced it even makes sense of the figures. The average age of pregnancy is currently 29.4 years (and that’s rather older than it used to be). Yet women in their 40s experience a greater gender gap than women in their 20s, 30s, or 50s.
If such a lack were to exist (and I don’t see that even your original figures support it), it might reflect a growth in credentialism (linked to rising numbers in higher education), rather than the disappearance of discrimination. One would have to follow the young people through their careers.
Tim, I’m not sure what you’re saying here. Are you really claiming you do not regard a 4 to 5% difference in mean wages (for 18 to 29-year-olds) as a gender gap in pay among younger people? Exactly how large a difference do you think constitutes a gap? Or are you conflating the question of whether there is a gap in pay with the problematics of why there is a gap in pay?
I absolutely agree that there is less discrimination then there used to be (which wouldn’t be difficult). As I hope you’re aware, many studies of the pay gap attempt to attempt to isolate the contributions of different factors to the gap. Usually in these studies, once direct factors other than gender are excluded, there remains a significant unexplained slice of the pay gap. (I know Farrell thinks he has managed to reduce this slice to nothing, but I wasn’t impressed by his op-ed into wanting to read his men’s rights polemic packaged as a women’s self-help book.) See, for example, Wendy Olsen and Sylvia Walby, Modelling Gender Pay Gaps (Equal Opportunities Commission, 2004) [PDF, 3.8 MB]. Olsen and Walby also astutely point out that gender influences a lot of the controlled factors, and thus there can be substantial indirect discrimination.
First, Ampersand is a he. Second, I wasn’t referring you to Ampersand’s wage gap series (I assume you meant posts plural) as a counterpoint to your statistical argument. Like you say, he’s discussing American figures (for the most part). But Ampersand provides a useful collection of feminist and pro-feminist thinking on the wage gap. He offers arguments which (I hope) would make you at least a little less triumphalist about some of your explanations of the wage gap (because your objections are often, in fact, part of feminist theories about why the pay gap exists and the basis of policy aiming to narrow it). But more importantly I’m trying to get you to realise that the EOC and feminist critics of the pay gap generally are not trying to argue there is a huge pay gap between men and women doing the very ‘same job’, narrowly defined. That’s not the primary focus of debate. If you go to the EOC website, for instance, they make clear that they are mainly talking about work of equal value.
So do I. But I believe I differ from you on three subsidiary points. First, I suspect women are over-penalized compared to men for absence from the labour force. Second, I think the bare fact that women are absent from the labour force for family care and not properly compensated for such work is a severe defect in our society (and with free market economic systems as they have traditionally been conceived). Third, it’s clear that women who do not have children also suffer from discrimination because they might have children. And I don’t need government statistics to show that. I can just take a wander over to your blog to find the following comment from the ever-charming ‘Rob Read’:
Just imagine the long-term impact of such practices on female earning power!
But back to you:
Actually, people are already trying to study this question. But even if it can be shown that a modern influx of women into a profession does not depress its wages, that wouldn’t mean there isn’t a gender gap. For one thing, the wages within the profession could polarize to leave women at the bottom and men at the top, while leaving average wages for the profession unaffected. But for another thing, historical discrimination can persist as a tradition of low pay. For instance, Paula England, Paul Allison, Yuxiao Wu, and Mary Ross, ‘Does Bad Pay Cause Occupations to Feminize, Does Feminization Reduce Pay, and How Can We Tell with Longitudinal Data?’ (presented at the annual meeting of the American Sociological Association, August 2004, San Francisco) [PDF, 274 kB] attempts to assess whether women-dominated occupations are low paid because women get low pay or because women get low paying jobs. They conclude that in the US between 1983 and 2001, falling wages in a given occupation did not cause feminisation, and feminisation did not lead to a fall in wages. They do not accept that this validates the theory that low wages in female-dominated jobs are balanced by non-pecuniary benefits, a possibility I think you raised. Rather they conjecture that early discrimination has been frozen into existing female-dominated occupations by institutional inertia and warn that ‘Without active policy intervention, the relative wages of female occupations will never reach parity with those of comparable — but different — male jobs’ (p. 31).
January 6th, 2006 at 11.24 am
My major complaint, and the reason I went looking for more statistics, was the way the EOC presented their figures a couple of weeks ago, leading to innumerable people (David Cameron included) stating that there was a 40% gender gap. Given the fact that the EOC deliberately excluded male part time wages from its calculations I regard this as casuistry at best.
“Second, I think the bare fact that women are absent from the labour force for family care and not properly compensated for such work is a severe defect in our society (and with free market economic systems as they have traditionally been conceived). Third, it’s clear that women who do not have children also suffer from discrimination because they might have children. And I don’t need government statistics to show that. I can just take a wander over to your blog to find the following comment from the ever-charming ‘Rob Read’:”
And that’s the point where I would be in vehement disagreement. In a free market “proper compensation” is, by definition, what people get. If you decide you want unfree markets all fine and well but you’ll also have to put up with the known effects of such intervention. Such as, the third point, that by forcing a business to pay women more than they receive in value from such employment you make the employment of women less likely.
I don’t even mind you arguing that this is what you desire. I just want to point out that you can’t have one without the other.
January 6th, 2006 at 12.29 pm
Tim,
I know your view on the EOC presentation of figures. I disagree, and may try to present a full argument as to why in a later post.
But (and I’m sorry to press you on this) do you now agree that there is a gender gap in pay, even for younger people?
By what definition? Would that still be true if I’d said ‘just compensation’?
Do you think social pathway dependency can distort the freedom of real markets? For instance, do you think social expectations that women do family care for free distort the market?
Well, that depends on the rest of your policy, e.g. whether you police hiring effectively, whether you compensate the employers with tax breaks, etc. But, moreover, studies show over and over again that you cannot explain away the gender gap with reference to absence from the job (see Olsen and Walby for just one example). Also, a lot of the equal pay debate is over underemployment as much as maternity leave, because there aren’t enough jobs with flexible working hours. Prima facie, this costs businesses too, since they’re effectively getting less productivity out of their female workers than they could. (And no, I don’t buy the argument that businesses are necessarily doing what’s best for them. They are as fallible as individuals, possibly more so.)
January 27th, 2006 at 1.07 pm
[…] According to Worstall, a gap in gross mean hourly earnings of 4.4% for 18 to 21 year olds approximates zero! […]