There’s a new paper out there looking at the progress on gender equality around the world. That paper, in the Journal of African Development, looks at three essential points. How much of the gender gap in education is being closed, how much of it in employment (and thus pay and income) and how much in political representation? Note that this is a global study so the numbers are going to be rather behind those of the UK and US. The way the paper is being read is that increased female access to education does not lead to greater gender equality of the other two. Yet this isn’t actually what the paper, properly understood, is saying at all. The important concept we have to add is of age cohorts. And once we do that then we can see that what we’re really being told is that greater access to education does indeed aid in closing those other two gaps. The point is though that this happens by age cohort: exactly the same as has happened in the UK and US in recent decades. There’s much greater gender inequality among 50 and 60 years olds in our own societies than there is among 25 year olds. And no, this isn’t solely the result of child bearing and rearing. It really was true that women had fewer career choices and less access to education 40 years ago than they do now. And some (please note, some) of that difference is reflected in the positions of those who were educated or not 40 years ago.

Here’s the Christian Science Monitor for example:

Women worldwide are closer to closing the global education gap with men, a years-long trend. Yet that progress hasn’t translated to women accessing equal employment opportunities, according to a new study.

For decades, international organizations such as the United Nations and the World Bank have been strong advocates of education, framing it as key to decreasing global inequality and eradicating poverty, as high levels of education are expected to raise income levels.

But a new study, published Friday in the Journal of African Development, noted that while women have attained 91 percent of the amount of education men have, measured by years in school, they lag behind in employment and political representation: 70 percent and 26 percent, respectively.

Lag behind? Most certainly. But that’s not at all the same thing as stating that the increased access to eduction isn’t working, or isn’t closing the gap. From the paper itself:

Employment Rate Ratios
Labor force participation rates measure the proportion of a country’s working-age population
that is active in the productive sphere12 of the economy, either by working for pay or looking
for work. Because the desire for paid work is not always fulfilled, it is useful to instead consider
employment-to-population ratios for those 15 and older. Caution should be used in making
inferences about well-being from these data since the definition of employment is broad.
Specifically, persons who have performed any work at all in the reference period for pay (of
any kind) or profit, or who were temporarily absent from a job for reasons of illness, parental
leave, holiday, training, or industrial dispute are counted as employed. This implies that the
economic effect of employment in terms of access to a livelihood varies widely, depending on
pay, hours of work, volatility of income, and other forms of non-wage compensation.
Figure 4 provides data on F/M employment-to-population rates for 177 countries for
1991 to 2010. As the data demonstrate, gender gaps are closing. The global ratio of F/M
employment rates rose from 0.62 in 1990 to 0.70 in 2010. In Panel A, the left tail of the
distribution of the ratio of F/M employment rates has shifted to the right. The lowest F/M
employment ratio in 1990 was a mere 9.8 percent (in Jordan). By 2010, the lowest ratio was
14.8 percent .

It’s getting better: the educational access has very definitely improved and so has access to work. Is there more to be done? Most assuredly. But we do now have to consider this age cohort business. Recall that this is all globally: so we are talking about some areas of the world where traditionally girls have not even received primary education, let along tertiary (as a purely personal note the elderly lady I bought this cottage from in rural Portugal is, even today, profoundly illiterate and innumerate, to the extent that the contract was signed with a thumb print and she was not able to tell me the number of the post box. 60 and 70 years ago, even in parts of Europe, it was not unusual out in the really rural areas that girls did not receive any education at all). So, that tradition has been coming to and end: as it darn well should have been doing. What’s the impact of that? Well, those who went through their childhood without being educated at all 40 years ago are not going to suddenly become educated because their female children and grandchildren are now being educated. It takes time for the first and second generations of educated children to move into the world of work and so on. The age cohorts really do move through society as cohorts. If we want to be brutal about it (and my mother insists, with 30 years experience of teaching this subject, that trying to teach literacy to an illiterate adult is something that just doesn’t work very well) those who are older now have had their chances ruined by their lack of education. That the newer generations are not being so ruined is a cause for celebration.
However, what we cannot do is look purely at the total population and then use that as a measure of how well greater female education is changing the society. We must look at the effects on the age cohorts. In the UK and in certain US urban centres young women have greater education than the young men around them. They also earn higher wages. Those who were offered a more limited tertiary education 30 years ago do not. So, should we measure the success or failure of the greater educational opportunities on the effects on those educated 30 years ago? Or upon those who received the greater opportunities? Should we be considering age cohorts or not? The answer, quite obviously, is that we can only measure the effects of that greater education upon those age groups that actually received it.

In other words that greater access to education is working: we can see that the general societal numbers are improving, it’s just that they’re going to improve slowly as that cohort which got the extra ages.

I would also say that there’s something rather odd about using female representation in industrial employment as the correct measure:

Therefore the bulk of the analysis is confined to an examination of gender differences in
three variables: 1) the ratio of female/male employment-to-population rates, 15 and older;
2) the ratio of male/female unemployment rates; and 3) the share of females and males
employed in the industrial sector to capture gender job segregation in the productive sector