Worldwide, one in five children of upper-primary-school age remain out of school. Girls in developing countries are disproportionately affected, with a quarter of them not completing primary school. In dealing with issues of gender inequality, some academics and policymakers have focused on the issue of menstruation as a reason why policymakers should (or should not) invest in sanitation-related interventions in schools (e.g., Birdthistle et al. 2011, Oster and Thornton 2011, Grant et al. 2013). However, an overly narrow focus on a girl’s few menstrual days may obscure the larger, every-day issues of health, privacy, and safety facing both girls and boys.
In my paper (“Sanitation and Education”), I use a change in national Indian policy to examine how improving the health, privacy, and safety of the school environment through sanitation influences the educational decisions of children across different ages. Specifically, I evaluate the educational impact of a large school-latrine-construction initiative using new annual administrative data on approximately 140 thousand Indian schools. To try to understand the mechanisms behind any impacts, I explore differential impacts by student sex and age.
The program
In 2003, the Government of India took on the issue of eradicating open defecation by providing substantial financial resources for schools to provide sanitation facilities. This initiative, called the School Sanitation and Hygiene Education program (SSHE), was bolstered by the Millennium Development Goals, in which it was suggested that school sanitation could help address several of their goals, including increasing school participation, promoting gender equality in education, combating disease, and ensuring environmental sustainability. Latrine construction was the main component of SSHE, which also included some hygiene education and small-scale investments (such as providing a bucket for water). The Ministry of Drinking Water and Sanitation generally managed the program, so there was no explicit tradeoff in other educational inputs. Due to resource constraints, the program was rolled out gradually over time.
A primary empirical challenge in estimating a link between sanitation and education is that schools with latrines may differ systematically from school without latrines. To overcome potential selection bias in the cross-section, I compare schools that receive latrines in 2003 to similar schools in the same district that do not receive latrines in 2003 using a differences-in-differences empirical strategy.
Estimated impacts on enrollment and achievement
Why do we see these increases? (Mechanisms)
To explore the mechanisms behind these impacts, I consider differential impacts by latrine type. Some schools received resources to build unisex latrines, whereas other schools received resources to build sex-specific latrines. These estimates will naturally be more suggestive because of selection bias concerns in which schools receive latrines of different types. I find that unisex latrines are mostly sufficient at younger ages with relatively smaller additional gains from sex-specific latrines, which may reflect important impacts through child health. At older ages, however, separate latrines become crucial. Pubescent-age girls benefit little from unisex latrines, and their enrollment increases substantially after the construction of separate sex-specific latrines, which suggests that privacy and safety concerns may be central to older girls’ decision-making. Understanding these differential impacts not only helps illustrate potential mechanisms, but is also of practical importance in deciding where to direct scarce resources: it is important to understand when unisex latrines are sufficient and when separate, sex-specific latrines are needed.
As an exercise to help understand the mechanisms further, I look at the impact of latrine construction on the share of female teachers, as one reason female teachers may be absent from school is due to the absence of sanitation facilities. I also explore whether enrollment effects are larger in schools with more female teachers initially. I find that a latrine increases the share of female teachers, which may lead to benefits for female students.
External Validity
A real advantage to having a large-scale setting with a large administrative dataset is that it allows me to explore the external validity of the estimates. In particular, I examine the geographic heterogeneity in the effects across districts with different cultural norms and income levels. The impact of a latrine is similar across the 269 districts in my sample, despite the significant heterogeneity across these states. There is suggestive evidence that the impact of school latrines is larger in areas with greater parity; however, the impact is still substantial in districts with lower gender parity. As an illustration of the geographic scope of my data, the variation across sample districts with average income at the 10th and 90th percentiles is comparable to the variation across countries between the 5th and 25th percentiles of the world income distribution (e.g., Rwanda and Nepal vs. Georgia and Ukraine).
Conclusion
Sanitation is a significant policy issue with substantial resources being put into it. In 2012 alone, UNICEF spent $380 million on the issue. Something that is this big of a deal deserves a large-scale evaluation to really influence what people are doing, but there has not been a large-scale evaluation of school sanitation.
Many schools worldwide lack sanitation facilities, but there is a growing resolve from governments and NGOs to invest in school-latrine-construction initiatives. Amidst the current focus on improving infrastructure in developing countries, it is important to remember basic needs. On the basis of empirical estimates from India's national initiative, efforts to improve sanitation worldwide might be expanded and re-directed to increase their impact.
Overall, these results illustrate the varying importance of health, privacy, and safety for children of different sexes and ages, which is useful to understand when trying to allocate scarce resources. While there are many deep roots to problems of gender inequality, improving school sanitation is one opportunity to increase gender equality for pubescent-age girls. Indeed, aspects of these results suggest that educational decisions of pubescent-age girls are influenced by physical, emotional, and societal changes that happen with the onset of puberty. However, a broader view is necessary to include the factors that influence the educational decisions of pubescent-age boys and younger children. The impacts of school sanitation on children, regardless of sex or age, suggest a central role for improving health, privacy, and safety in increasing educational attainment worldwide.
Anjali Adukia is a doctoral student at Harvard University. Her research interests primarily focus on issues related to education, children, and poverty-alleviation questions in the context of developing countries.
In my paper (“Sanitation and Education”), I use a change in national Indian policy to examine how improving the health, privacy, and safety of the school environment through sanitation influences the educational decisions of children across different ages. Specifically, I evaluate the educational impact of a large school-latrine-construction initiative using new annual administrative data on approximately 140 thousand Indian schools. To try to understand the mechanisms behind any impacts, I explore differential impacts by student sex and age.
The program
In 2003, the Government of India took on the issue of eradicating open defecation by providing substantial financial resources for schools to provide sanitation facilities. This initiative, called the School Sanitation and Hygiene Education program (SSHE), was bolstered by the Millennium Development Goals, in which it was suggested that school sanitation could help address several of their goals, including increasing school participation, promoting gender equality in education, combating disease, and ensuring environmental sustainability. Latrine construction was the main component of SSHE, which also included some hygiene education and small-scale investments (such as providing a bucket for water). The Ministry of Drinking Water and Sanitation generally managed the program, so there was no explicit tradeoff in other educational inputs. Due to resource constraints, the program was rolled out gradually over time.
A primary empirical challenge in estimating a link between sanitation and education is that schools with latrines may differ systematically from school without latrines. To overcome potential selection bias in the cross-section, I compare schools that receive latrines in 2003 to similar schools in the same district that do not receive latrines in 2003 using a differences-in-differences empirical strategy.
Estimated impacts on enrollment and achievement
- School latrines increase enrollment and lower dropout for all students, not just pubescent-age girls. For example, enrollment increases by 12% in primary schools (1st-5th grades) and 8% in upper-primary schools (6th-8th grades).
- These changes are also reflected in the increased number of students who appear for and pass the middle school board exam. Impacts on student achievement reflect the net effect of the introduction of a school latrine: new students may perform well on the exams, previous students may perform better due to improvements in the school environment, and/or previous students may perform worse due to overcrowding.
- A latrine generally increases female enrollment more than male enrollment.
- Inconsistent with a narrow focus on menstruation, younger girls and boys experience larger benefits than older children from the presence of a latrine.
- While the impact of educational infrastructure interventions can fade-out over time, the enrollment impact of latrines is still present or even slightly stronger three years later. This is particularly striking, as the latrine-construction initiative did not specifically target latrine maintenance.
Why do we see these increases? (Mechanisms)
To explore the mechanisms behind these impacts, I consider differential impacts by latrine type. Some schools received resources to build unisex latrines, whereas other schools received resources to build sex-specific latrines. These estimates will naturally be more suggestive because of selection bias concerns in which schools receive latrines of different types. I find that unisex latrines are mostly sufficient at younger ages with relatively smaller additional gains from sex-specific latrines, which may reflect important impacts through child health. At older ages, however, separate latrines become crucial. Pubescent-age girls benefit little from unisex latrines, and their enrollment increases substantially after the construction of separate sex-specific latrines, which suggests that privacy and safety concerns may be central to older girls’ decision-making. Understanding these differential impacts not only helps illustrate potential mechanisms, but is also of practical importance in deciding where to direct scarce resources: it is important to understand when unisex latrines are sufficient and when separate, sex-specific latrines are needed.
As an exercise to help understand the mechanisms further, I look at the impact of latrine construction on the share of female teachers, as one reason female teachers may be absent from school is due to the absence of sanitation facilities. I also explore whether enrollment effects are larger in schools with more female teachers initially. I find that a latrine increases the share of female teachers, which may lead to benefits for female students.
External Validity
A real advantage to having a large-scale setting with a large administrative dataset is that it allows me to explore the external validity of the estimates. In particular, I examine the geographic heterogeneity in the effects across districts with different cultural norms and income levels. The impact of a latrine is similar across the 269 districts in my sample, despite the significant heterogeneity across these states. There is suggestive evidence that the impact of school latrines is larger in areas with greater parity; however, the impact is still substantial in districts with lower gender parity. As an illustration of the geographic scope of my data, the variation across sample districts with average income at the 10th and 90th percentiles is comparable to the variation across countries between the 5th and 25th percentiles of the world income distribution (e.g., Rwanda and Nepal vs. Georgia and Ukraine).
Conclusion
Sanitation is a significant policy issue with substantial resources being put into it. In 2012 alone, UNICEF spent $380 million on the issue. Something that is this big of a deal deserves a large-scale evaluation to really influence what people are doing, but there has not been a large-scale evaluation of school sanitation.
Many schools worldwide lack sanitation facilities, but there is a growing resolve from governments and NGOs to invest in school-latrine-construction initiatives. Amidst the current focus on improving infrastructure in developing countries, it is important to remember basic needs. On the basis of empirical estimates from India's national initiative, efforts to improve sanitation worldwide might be expanded and re-directed to increase their impact.
Overall, these results illustrate the varying importance of health, privacy, and safety for children of different sexes and ages, which is useful to understand when trying to allocate scarce resources. While there are many deep roots to problems of gender inequality, improving school sanitation is one opportunity to increase gender equality for pubescent-age girls. Indeed, aspects of these results suggest that educational decisions of pubescent-age girls are influenced by physical, emotional, and societal changes that happen with the onset of puberty. However, a broader view is necessary to include the factors that influence the educational decisions of pubescent-age boys and younger children. The impacts of school sanitation on children, regardless of sex or age, suggest a central role for improving health, privacy, and safety in increasing educational attainment worldwide.
Anjali Adukia is a doctoral student at Harvard University. Her research interests primarily focus on issues related to education, children, and poverty-alleviation questions in the context of developing countries.
Join the Conversation