Human Development

How to keep more girls in school? Lessons from Bangladesh

  • Blog Post Date 24 September, 2012
  • Articles
  • Print Page

For years developing countries have been trying to increase parents’ incentives to send their children, particularly girls, to school and keep them there. This column looks at the success of Bangladesh, where the number of girls in school now exceeds the number of boys. It argues that money talks – but it’s the money that educated children will earn once they leave school that talks loudest.

Over the past decade the school enrolment rate of Indian children has risen substantially, both at primary and secondary school levels. This has been particularly true for girls, whose share in total enrolment has increased from 47.52% in academic year 2004-2005 to 48.46% in 2009-2010 (Government of India 2009). However, as recently reported on this website (see Farzana Afridi’s column on the design of school subsidy programmes, Afridi 2012), there continue to be gaps in both enrolment and attendance: the gross enrolment ratio for secondary education is only 60% (UNESCO 2007), and the attendance rate in primary school is just under 70% (Government of India 2011).

Although traditionally many programmes targeting education have focused on simply improving school infrastructure, under the theory of ‘build it and they will come,’ in recent years, a growing number of programmes worldwide (including in India) have attempted to directly increase school enrolments through ‘conditional incentives’ – rewards dependent on school attendance. Some of these seek to target the direct costs of schooling, offering subsidies for uniforms, school supplies, and tuition (in areas where education is not provided for free). These programmes assume that parents want to invest in their children’s education, but are unable to do so because they cannot afford it.

Randomised evaluations in Kenya have shown positive effects from these programmes: in one programme, attendance increased 6% after school-aged children were provided with a free uniform, with a larger increase (13%) for those who had not had a uniform prior to their participation in the programme (see for instance Evans et al. 2008). Other conditionality-based programmes focus their efforts on changing the incentives parents have to send their children to school or keep them at home. Implicit in the design of such programmes is the idea that the ‘opportunity costs’ of schooling – the best alternative forgone – in terms of lost jobs and income are prohibitively high. The programme offsets these costs by providing an immediate cash rewards for keeping the child in school. In Mexico’s Oportunidades programme, for example, low-income families receive cash transfers that can be worth almost a quarter of their income (Fiszbein and Schady 2009), and the grants increase in size as the children age, to reflect the increased opportunity cost of keeping an older child out of work (Rawlings and Rubio 2005).

Yet, while many of these programmes have had considerable success in improving enrolments and attendance, reducing the costs of attending – either the direct costs or the indirect opportunity costs – does not address the long-term value that parents place on education and therefore their willingness to invest in education for their children. If the job prospects in the area are as dismal for those with an education as they are for those without one (or if just basic literacy skills will suffice for most available jobs), it is no wonder that parents – even ones who could afford to do so – choose not to send their children to school.

In a new study using data from Bangladesh, Rachel Heath and I examine the impact of job availability on parents’ decision to invest in their children’s education. Bangladesh provides a unique case study, as it is one of the few developing countries to have met the third Millennium Development goal of achieving gender parity in education. Numerous studies have attributed this impressive feat to the introduction of the Female Stipend Program (FSP) in 1994, a conditional cash transfer programme that gives a monthly stipend to female secondary school students contingent on maintaining a minimum attendance rate and GPA, and remaining unmarried. However, little empirical data exists on the FSP, and only two other researchers that we know of have attempted to rigorously evaluate the impacts of the FSP (see Nobuhiko 2001 and Khandker et al. 2003).

Just prior to this policy change, the Bangladeshi economy had begun to undergo a huge growth in its export sector, particularly in the area of ready-made garments. This new sector has had its largest impact on the availability of jobs for women: approximately 80% of garment workers are female, and garment jobs represent the best – and sometimes the only – option for women to work outside the home. Such jobs also provide a strong incentive for girls to stay in school, as they typically require literacy, numeracy, cognitive skills to be able to work in a team-production environment, and some ability to read and identify clothing labels in English.

This coming together of economic and social policy change provides a unique opportunity to examine the relative importance of each in increasing school enrolment rates. To isolate the effect of the growth in garment sector job availability, we use a ‘triple difference’ strategy. That is, we look at the difference in enrolment rates between children in villages within commuting distance of a garment factory and those which are not, the change over time as the garment sector grew and more jobs became available, and the difference between girls and boys within the same family, as access to garment jobs disproportionately affects females. Similarly, to estimate the effects of the FSP, we look at the difference in enrolments between those who have been in school for at least six years (and are therefore eligible for the programme) and those who have not, the difference before and after 1994 (the year of programme implementation) and, again, the difference between girls and boys in the same family, as the programme was only available to girls.

We find that, for girls aged 5-10, a 10% growth in garment sector employment increases the probability that the girl will be enrolled in school by 1%, and this is statistically significant. Meanwhile, availability of garment jobs actually decreases the likelihood of enrolment for girls over 16, presumably because at this age they drop out in order to take the jobs. This is not the case for boys in the same age groups. Although this figure may seem small, it translates into a 27 percentage point gain in overall girls’ enrolment over the last 30 years as the government industry flourished. A rough back-of-the-envelope calculation suggests that this increase in schooling demand associated with garment sector growth accounts for nearly all the gains in enrolment in villages close to garment factories, and between 20-25% of the gains in enrolment nation-wide.

In contrast, when we look at the results of the Female Stipend Program, we fail to find any statistically significant positive results. In fact, since the growth in female enrolment rates pre-dates the introduction of the FSP, it seems likely that previous evaluations attributing this growth to the FSP may have conflated the programme effects with pre-existing trends, such as those caused by the growth in the garment sector.

Importantly, the effects of garment industry jobs go beyond just increased education for young women. We find that, for a 10% increase in garment jobs, a girl’s propensity to be married before age 18 decreases by 7.3%. Her likelihood of having a child before age 18 also decreases, although these results are not statistically significant at conventional levels. These changes are of interest because other research has shown that early marriage and childbirth can have negative effects on child development and women’s well being (See for example Jensen and Thornton 2003 and Hotz, et al.2005).

Lessons for India and beyond

These results have important implications for education policy in developing countries worldwide. They suggest that, while improving the quality and quantity of educational inputs may be important, parents’ willingness to invest in their children’s education may be an even stronger determinant of child school enrolment. And rational households would only be willing to make this investment if there are positive returns to schooling, i.e., if jobs that require education and pay better wages are actually available. Thus education policy is tied to trade and industrial policy that creates skilled job opportunities. In the case of Bangladesh, a high-performing export sector that created positive returns to educating girls was able to readily overcome any traditional ‘cultural’ biases against educating girls. In summary, ‘money talks’. And now girls’ school enrolment rates in Bangladesh exceed that of boys.

Further Reading

  • Afridi, F (2012), “Food for thought: On the design of school subsidy programmes”,, 6 August 2012 Evans, D, M Kremer, and M Ngatia (2008), “The Impact of Distributing School Uniforms on Children´s Education in Kenya”, Unpublished Draft.
  • Fiszbein, A and N R Schady (eds.) (2009), Conditional Cash Transfers: Reducing Present and Future Poverty, World Bank Publications.
  • Fuwa, Nobuhiko (2001), The Net Impact of the Female Secondary School Stipend Program in Bangladesh, University Library of Munich, MPRA Paper. No. 23402.
  • Government of India (2007), “Study of Students’ Attendance in Primary and Upper Secondary Schools, Abridged Report”, Research, Evaluation and Studies Unit Technical Support Group.
  • Government of India (2009). “Elementary Education in India: Progress Toward UEE”, District Information System for Education, Flash Statistics.
  • Heath, R and A M Mobarak (2012), “Does Demand or Supply Constrain Investments in Education? Evidence from Garment Sector Jobs in Bangladesh”, Working Paper.
  • Hotz, VJ, S McElroy and S Sanders (2005), “Teenage Childbearing and its Life Cycle Consequences: Exploiting a Natural Experiment”, The Journal of Human Resources, 40(3):683-715.
  • Jensen, R and R Thornton (2003), “The Consequences of Early Marriage in the Developing World”, Oxfam Journal of Gender and Development, 11(2):9-19.
  • Jensen, R (2010a), “The (Perceived) Returns to Education and the Demand for Schooling”, Quarterly Journal of Economics, 125(2):515-548.
  • Jensen, R (2010b), Economic Opportunities and Gender Differences in Human Capital: Experimental Evidence from India, NBER Working Paper Series, No. 16021.
  • Khandker, SR, MM Pitt, and, F Nobuhiko (2003), Subsidy to Promote Girls’ Secondary Education: The Female Stipend Program in Bangladesh, MPRA Paper. No.23688.
  • New York Times (2010). “Garment Factories, Changing Women’s Roles in Poor Countries”, 21 July.
  • Rawlings, LB and GM Rubio (2005), Evaluating the Impact of Conditional Cash Transfer Programs, The World Bank.
  • The Daily Star, Bangladesh (2012). “RMG growth boosts female school enrolment: Two economists tell the success story”, 16 July.
  • The Straits Times, Singapore (2012), “Garment Sector Sparks Change in Bangladesh”, 20 August.
  • UNESCO (2011), “The Hidden Crisis: Armed Conflict and Education”, UNESCO Publishing.
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter