Module 7: Effective Education

Student Attendance

Although tremendous improvements have been made since 2000, the United Nations estimated that in 2018 258 million children and adolescents, or one-fifth of the population in that age group, were not attending school.(1)  Impoverished children may not attend school due to gender discrimination, a lack of income, the high cost of school fees and supplies, individual health problems, or even simply menstruation.  With the diversity of barriers to attendance come a plethora of potential solutions.  To increase student attendance, researchers have tested a variety of interventions, some of which seem obvious but fail, and others which seem irrelevant but succeed. 

Problem With Assumptions

For years, organizations and governments have been dumping money into schools believing that improvements in teaching, textbooks, and/or materials would increase attendance.  However, researchers at MIT’s Abdul Latif Jameel Poverty Action Lab (J-PAL) have shown that increasing the number of textbooks, flipcharts, or teachers has little effect on student attendance rates.  Better resources and equipment in schools are not enough to incentivize students to attend or parents to send their children to school, though these improvements may provide those students attending school with a better education and better opportunities in the future.(2)

Policy makers have also promoted a belief that girls miss school during menstruation due to a lack of sanitary products.  This phenomenon could serve as an easy explanation for why female school attendance lags behind that of males in developing countries.  While the World Bank states that girls may miss up to 20% of the school days per year, it is assuming that girls always miss school during their period.  Because of the promotion of this idea, Proctor & Gamble has donated $5 million to provide sanitary products and puberty education to students in the developing world with the aim of increasing female school attendance.  This assumption has one supporting study.  The Scott et al. (2009) study found that after the provision of free sanitary pads to girls in Ghana, the days of school missed was cut in half, resulting in a recovery of about one week of school per term.  However, the authors acknowledge the limitations of the study, including its short duration (Winter/Spring school term), small number of sites, and relatively small sample size, as well as the novelty effect of the intervention in the rural areas.  They aptly note, “it is essential that further study over a longer period of time be done before policy decisions committing substantial funding, especially from poor governments, occurs”.(3)   A more recent study by Oster & Thornton (2010) was conducted over a longer period of time (15 months), and it found conflicting results that suggested that funding for sanitary pads may not be the most appropriate allocation of resources to boost female student attendance.(4)

Researchers at the University of Chicago executed a randomized evaluation in Nepal that debunked the assumptions that 1) girls miss significant amounts of school during their periods and that 2) absences are due to a lack of access to sanitary products.  According to this study, girls attend school 85.7% of the time on non-period days and 83.0% of the time on the days of their menstruation.  Far below the estimates of the World Bank, the 2.7 percentage point difference indicates that menstruation accounts for an insignificant proportion of female absenteeism.  Overall, menstruation accounted for girls in this study missing only one third of a day per year due to their period.  Moreover, when half of the girls in the study were given sanitary products such as menstrual cups, no significant change was observed in attendance rates.  The most reported reason for missing school during menstruation was cramps, which 43.8% of the girls noted.  Only 20% of girls reported poor sanitary products and mobility restriction as their reason for absenteeism.(5)  Although it is a noble cause to provide girls with better sanitary products, foundations must realize that it is not the most effective way to boost female school attendance and may not, in fact, accomplish that goal at all. Both of the examples listed above indicate that explanations with only anecdotal evidence may not represent reality; therefore, these assumptions do not suffice as bases for the widespread implementation of a certain intervention. 

In fact, less obvious and less cited factors may contribute more to student absenteeism than would be expected.  For example, providing iron supplements and deworming pills have proven to be cost-effective means to drastically raise school attendance rates even though worms and anemia are rarely reported as reasons for absenteeism by parents and children.

Case Studies in Effective Interventions

Deworming at School in Kenya

Intestinal worms do not show up on the list of barriers to school attendance above because they are not an obvious culprit.  However, 400 million school-age children suffer from chronic infection with intestinal worms, the symptoms of which include diarrhea, abdominal pain, anemia, and lethargy.  Although a cheap and highly effective oral medication exists to treat this infection, only 10% of at-risk children are treated.  Considering these conditions, researchers of J-PAL designed an intervention in which 30,000 children in 75 primary schools in rural Kenya were treated for hookworm, whipworm, roundworm, and schistosomiasis in order to determine the effect of deworming on school attendance.  The study found that not only did the health program reduce the proportion of children with moderate-to-heavy infections by 23 percentage points on average, but it also decreased school absenteeism by 25%.  Interestingly, similar effects on education were seen in the U.S. in 1915 when intestinal worms were eradicated in the southern states.  This program was the most cost-effective intervention of all of those tested, resulting in 649 DALYs (Disability Adjusted Life Years) saved in the first phase of the program and costing US$5 per DALY saved.  This program resulted in 28.6 additional years of student attendance per $100 spent.  It is important to note that while the program led to improvements in students’ health and attendance rates, no impacts on test scores were observed. Thus, increased attendance does not necessarily correlate with higher test scores.(6)  However, the positive health effects were, of course, a benefit in themselves.

Information on Higher Wage Returns to Education in Madagascar

In this intervention, the effects of role models and of the presentation of statistics to school children’s parents were examined in 640 schools in Madagascar.  In the first group of schools, teachers presented to parents and children the average earnings associated with each level of education completed.  In the second group, a role model presented to parents and children his or her background, education level, and current earnings.  In the third group, role models presented the same information as above along with statistics concerning returns on education.  In the first “statistics only” group, test scores improved by 0.2 standard deviations.  By the end of the intervention, the attendance by students in schools presented with statistics was 3.5 percentage points higher than in the control group, and in the Dominican Republic, this presentation of financial returns reduced the number of dropouts by 7%.  Interestingly, in the third group, the presence of the role model negated the positive effects of the presentation of statistics.(7)  Since the intervention only required a meeting with parents and children, this intervention was highly cost-effective, gaining an additional 40 years of student attendance per $100 spent on the intervention.(8)

While increasing student attendance is the first step in achieving universal education, it does not necessarily ensure better opportunities for children.  Reaching 100% attendance of all children in primary schools would be a formidable output, but the outcomes of income generation and poverty alleviation are the end goals to be achieved.  Such outcomes must be measured in addition to attendance rates.  See Outcomes are Essential in Global Health for more information on the distinction between outputs and outcomes.  J-PAL acknowledges this complexity and writes: “It is assumed that if children go to school more often they will learn more.  But this isn’t necessarily the case.  Several programs which have raised participation, from providing worm medicine to free meals, show no evidence that children are learning more as a result.  Perhaps only a large increase in performance would show up as significant. Or perhaps other factors—such as better teaching methods or more regular attendance of teachers—are necessary for participation to raise test scores.”(9)  As mentioned, teacher absenteeism can significantly decrease both student attendance and performance and is a significant problem in many parts of the developing world.

Case Studies in Effective Interventions

Teacher Monitoring in India

In rural Udaipur, India, 44% of teachers are absent.  A study run by researchers at J-PAL aimed to increase school attendance in this region by monitoring teacher attendance with cameras.  In the study group, teachers were given cameras and were paid according to how many days they recorded their attendance through photos with the students.  Not only did teacher attendance increase from 58% in the control group to 79% in the intervention group but also the monitoring resulted in students being 62% more likely to be admitted to regular government schools.  Utilizing objective monitoring with incentives, this intervention resulted in 45 additional days of provider attendance per $100 spent.(10)

Merit-Based Girls’ Scholarships in Kenya

Even more successful than the camera-monitoring was providing merit-based girls’ scholarships and hiring extra contract teachers.  In Kenya, J-PAL implemented an intervention in select schools in which 6th grade girls who scored in the top 15% on government-administered tests earned a grant of US$6.40 to cover school fees, US$12.80 for school supplies, and public recognition in their communities. Overall, both the girls and boys scored better on tests, while student absenteeism decreased by 25% in one of the two districts studied.  Moreover, teacher attendance increased by 4.8 %age points.  This improvement may be due in part to the pressure exerted by parents on the teachers to help their children win scholarships.  This intervention resulted in 50 additional days of provider attendance per $100 spent.(11)  However, when teacher pay was dependent on student test scores in Kenya and India, teacher absenteeism rates did not change.  A rise in test scores was seen, but could be largely attributed to test preparation.(12)  Interestingly, rewarding the students rather than the teachers for better performance may be a more effective way of increasing teacher attendance.

Hiring Local Contract Teachers in Kenya 

An intervention in Kenya raised teacher attendance using a different hiring scheme.  Rather than trying to improve upon the existing educators alone, J-PAL hired civil service teachers to address the overcrowding of classrooms.  When classrooms were split between teachers, students were randomly assigned to either the civil service teacher or the contract teacher in one half of the intervention schools.  In the other half of the schools, students were assigned based on their level of preparedness, and the teachers were randomly assigned to either the higher- or lower-preparedness section.  Half of each group of schools was trained with additional local school committee oversight of the contract teacher.  Although the contract teachers were paid about one-quarter of the civil service teacher salary, they had lower absenteeism rates than civil service teachers, and their students scored higher on exams.  With the added oversight, program effectiveness was even greater.  The hiring of local contract teachers proved to be a cost-effective way to promote learning and decrease teacher absenteeism.  The number of additional days of provider attendance per $100 spent is impossible to calculate, as money was actually saved in the program by hiring contract teachers as compared to hiring more civil service teachers.(13)

These and other studies by J-PAL demonstrate that “impersonally administered and direct incentives for attendance are the most effective at increasing service provider attendance…Attempts to increase accountability to the poor were successful in cases where monitors had credible authority which the service providers respected…”  In cases where the supervisors had full control over administering incentives, the programs were ineffective.  In Kenya, for example, when school providers were fully responsible for distributing the incentives, they distributed them irrespective of the provider’s attendance record.  On the other hand, in India, when community monitoring of teachers was implemented and the community had no control over hiring or firing the teachers, the program was also ineffectual. (14)  It appears that while community control may be necessary in some cases, it may prove detrimental in others.  More research is necessary to determine which schemes work best in certain contexts and cultures.  While some of the interventions tested did not improve teacher attendance at all, others raised teacher attendance rates as well as improved student test scores and/or graduation rates.

Remedial Education

Providing remedial education is another method used to improve reading and comprehension levels.  While some countries are making great strides in achieving universal student attendance, learning levels are lagging behind.  For example, the enrollment rate in Ghana is 95%; however, less than half of 3rd grade students have the minimum competency level in reading, writing, and basic arithmetic.  In order to improve the learning levels of students, researchers at Innovations for Poverty Action have implemented remedial education in schools in Kenya and India.  One intervention in India hired local young women, called balsakhis, who had completed at least secondary education to teach lower performing students basic math and language skills for two hours during the school day.  The children in the remedial education program were 2nd, 3rd, and 4th grade students who had not yet displayed the math and reading competencies taught in 1st and 2nd Grade.  Another program used volunteer teachers to provide this extra education outside of school.  The Kenya intervention hired contract workers to provide remedial education and its success is described above.(15)

The remedial education provided in the first program resulted in improvements on all math and verbal tests in a highly cost-effective manner.  For example, at the start of the second year of the program, only 5 to 6% of the students could add two-digit numbers.  At the end of the year, 51% of the students who were taught by balsakhis had this math competency, compared to 39% among those who were not taught by balsakhis. This intervention led to especially significant gains in competencies among marginalized and poorer students who are often the ones left behind.(16) 

It is important not only to increase student attendance so that all children have the opportunity to earn an education, but also to work toward the end goal of increasing student competencies, especially among those who stand to benefit the most from education.  Pouring money into teachers and materials for schools is futile unless it is accompanied by measures to ensure that students are able to take full advantage of those resources.   Hypotheses continue to be tested to ensure that funding for schools, books, and equipment are not wasted and that poor children are not put at further disadvantage due to teacher absenteeism and inadequate education. 

Footnotes

(1) United Nations. (2020). “Quality Education: Why it Matters.” https://www.un.org/sustainabledevelopment/wp-content/uploads/2017/02/4_Why-It-Matters-2020.pdf .

(2) “Student Attendance.” Abdul Latif Jameel Poverty Action Lab.  Massachusetts Institute of Technology, 2009. 

(3) Scott L., Sue D., Paul M., Catherine D., and Caitlin Ryus. “Impact of Providing Sanitary Pads to Poor Girls in Africa.” University of Oxford, November 2009.

(4) Oster, Emily and Rebecca Thornton. “Menstruation, Sanitary Products and School Attendance: Evidence from a Randomized Evaluation.”  American Economic Journal: Applied Economics 27 April, 2010.

(5) Ibid.

(6) “Mass Deworming: A Best Buy for Education and Health.” Abdul Latif Jameel Poverty Action Lab. Policy Briefcase No. 4.  Massachusetts Institute of Technology, October 2007. 

(7) Nguyen, Trang. “Information, Role Models and Perceived Returns to Education: Experimental Evidence from Madagascar.” Massachusetts Institute of Technology, 23 January 2008.

(8) “Student Attendance.” Abdul Latif Jameel Poverty Action Lab.  Massachusetts Institute of Technology, 2009. 

(9) “Fighting Poverty: What Works?” Abdul Latif Jameel Poverty Action Lab.  Issue One.  Massachusetts Institute of Technology, Fall 2005. 

(10) “Solving Absenteeism, Raising Test Scores.” Abdul Latif Jameel Poverty Action Lab. Policy Briefcase No. 6.  Massachusetts Institute of Technology, September 2008. 

(11) “Incentives to Learn: A Merit-Based Girls’ Scholarship Program in Kenya.” Abdul Latif Jameel Poverty Action Lab.  Massachusetts Institute of Technology, 2009. 

(12) “Teacher Attendance.” Abdul Latif Jameel Poverty Action Lab.  Massachusetts Institute of Technology, 2009. 

(13) “Peer Effects, Pupil-Teacher Ratios, and Teacher Incentives in Kenya.” Abdul Latif Jameel Poverty Action Lab.  Massachusetts Institute of Technology, 2009. 

(14)  “Teacher Attendance.” Abdul Latif Jameel Poverty Action Lab.  Massachusetts Institute of Technology, 2009. 

(15) “Remedial Education.” Innovations for Poverty Action, 2011.

(16) “Making Schools Work for Marginalized Children.” Abdul Latif Jameel Poverty Action Lab. Policy Briefcase No. 2.  Massachusetts Institute of Technology, November 2006.