Published on Development Impact

When it comes to female education, have we gotten it all backwards?

This page in:

To get children to attend school in developing countries, our approach has been primarily to assume that the schooling that is available is worth pursuing, meaning that the problem must be with some barrier to go to school despite a great desire to do so: perhaps the family cannot afford the costs of schooling; perhaps the schools are not good or too far; perhaps the children want to be in school but the parents prefer food today to educated daughter tomorrow; maybe people don’t know the value of schooling, etc. These ideas have been behind many popular development programs to improve education, such as building and improving schools, conditional cash transfers, information campaigns, etc. But, have we gotten this whole thing all backwards?

An alternative explanation is that most households know the value of schooling pretty well and they rationally decide to not send their daughters to school. This could be because the expected accumulation of this human capital is not rewarded for adult women in society – there are no good jobs for them or no good prospect of meaningfully participating in society, i.e. have autonomy and agency. If that is the case, other interventions/factors that increase the returns to schooling may be more effective in causing families to invest in their daughters than the types of interventions outlined above. And the evidence on this is finally accumulating and is becoming harder to overlook.

In the series of guest posts we ran at the end of last year by students in the job market (and junior faculty), there were two papers that suggested that changes in the availability of jobs for adult women can lead to changes in human capital investments in female children. One of these papers by Heath and Mobaraq (see Heath’s post on their paper here) argued that the rapid expansion in the garment industry in Bangladesh led to increases in girls’ schooling, which did not happen for boys. As the effects were also observed among households in which no adult had a garment industry job, the authors argue that the increase was not just due to income effects, but could also be due to the increased returns to education for young women or increased bargaining power for mothers.

In a second paper, Majlesi examines changes in demand for the labor of women of different ages in Mexico and finds that increases in demand for women 25 and older improves self-reported health of daughters and their chances of having completed vaccinations (see his post here). Again, no effect is found for boys. No effect is found for increased demand for younger women either, which the author argues to mean that the changes are happening not through a increased returns channel (or an increased income channel) but rather through increased bargaining power of mothers. However, as he is examining health outcomes rather than education, it is hard to tell whether families are not investing in older school-age girls more in response to the increased future earnings.

These two papers are working papers (and you know how I feel about those) that suggest that suitable jobs for women that are relatively safe and pay a decent wage may spur families to invest in their daughter’s schooling and close the gender gaps in education and aspirations for (and among) children – either through increased bargaining power of mothers or increased returns to schooling for the school-aged. The latest evidence in this line of work comes from a randomized controlled trial by Rob Jensen in India in a paper that is now forthcoming in the QJE.

Rob, ahead of most of us in running these experiments, went to India in 2003 and hired himself 8 recruiters who would go to villages that no other recruiters would visit (because they were too far from Delhi, hence not profitable for the recruiters) and advertise jobs in the Business Process Outsourcing industry to young women. These are mostly call center jobs, with some clerical and IT jobs in the mix and the employers in this rapidly growing new industry in India at the time preferred young female workers with secondary education. He finds that young women aged 15-21 in randomly selected treatment villages were more likely to obtain more schooling, pay for private post-school training in English or computers, while also less likely to get married and have children. Their aspirations also increased, with more of these women wanting to work after marriage and to have fewer children.

While these findings are really interesting, what is truly impressive about the paper is the analysis of the channels. Because Jensen set out “simply to test as cleanly as possible the effect of employment of the young woman herself, or the potential future employment (for girls),” he designed his experiment accordingly and thought about how he would systematically rule out each alternative explanation. He manages to do so convincingly and argues that the effects were not through increased wealth, increased bargaining power among mothers, or changes in time allocation or fertility among parents. Young women were simply responding to the availability of the new jobs (rather to information about the availability of these new jobs) and invested significant amounts in their human capital to be able to get these (or similar) jobs – now or in the near future.

Finally, same changes can also happen because girls (and their families) have something else to look forward to: meaningful participation in society, such as political participation. Markus scooped me earlier this week and described what happens when “women are in charge” when he discussed the case of reserved village leadership positions for women in India. In an article in Science Lori Beaman and co-authors find that the gender gap in adolescent educational attainment was erased in villages where a female leader was randomly assigned for two cycles. Because there are no changes in young women’s labor market opportunities, the authors think that this is mostly a “role model” effect: the aspirations of both the girls and their parents for those girls changed substantially.

So, if girls have something to look forward to, be it good jobs or a more fulfilling life with autonomy and agency, households are willing to invest in their education. In the case of India, the costs of the investments are non-negligible and the households were willing to make them without explicit accompanying interventions that helped them offset these costs. Perhaps parents (and societies) are more willing to change gender norms (and quickly, as Markus observed in his post on Tuesday) than we thought. Of course, households may still be constrained in other ways and the traditional schooling interventions described in the first paragraph above could still be effective counterparts to increased jobs for women or improvements in gender equity: after all, if there is not a sufficient number of women with skills to fill these positions, such policies will lead to failure and risk setting progress back. In both Bangladesh and Mexico, there were food for education or conditional cash transfer programs being implemented by the government while job opportunities for women were on the rise.

On the other hand, such investments in education without jobs or increased gender equity in society can also lead to frustrated aspirations for many. So, what can governments do so that there are more jobs available for women in the formal sector? These are areas of trade and industrial policy, in which I am out of my depth, but would like colleagues to chime in. The next World Development Report is on jobs and I have many colleagues currently working on the report: can the current approaches be tailored to ensure that new jobs are available for young women? This group is of particular concern, due to the fact that youth unemployment is high and women are lagging behind (see a recent post on this issue by Tamar Manuelyan Atinc here).

I say formal sector jobs above because of another concern I have: sextortion. A colleague of mine, Mary Hallward-Driemeier, who has written a report titled “Enterprising Women: Expanding Opportunities in Africa” (see a blog post by her here), has recently given a talk where she showed the audience data that quantifies the stories we all heard working in Africa: running a small business as a woman is also treacherous business. Many say they had to exchange sex with a person of authority (person in charge of a permit, border police, etc.) or a supplier in order to be able to go about their business. Many people think microfinance or similar interventions to help poor women in Africa, but in the current environment the emerging small enterprises may not only have small profit margins and little room to grow, but also can be damaging to the dignity, mental health, and physical health of these women. Jobs in sectors that are somewhat regulated, that require more brain than brawn, may provide safer and more acceptable alternatives to running one’s own business – until at least such time when we’ve figured out how to make the latter more attractive to women.

Please share your thoughts in the Comments section…


Berk Özler

Lead Economist, Development Research Group, World Bank

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000