Over the past decade many developing countries have expanded primary school access, energized by initiatives such as the United Nations Millennium Development Goals, which call for achieving universal primary education by 2015. However, improvements in school enrollment have not always translated into actual improvements in skills and learning for students. Many schools struggle with high dropout rates, possibly due to the direct costs of fees and materials and opportunity costs families face with fewer individuals in the workforce or providing help at home. Evidence from programs implemented in Mexico, Honduras, and Nicaragua, among others, suggest that a cash transfer program, conditional on school-aged children attending school, could help reduce dropout rates. However, there is no evidence to date on the degree of monitoring of school attendance that is necessary. Nor is there evidence on whether the conditionality itself is necessary to ensure increased attendance or whether the CCT increases income and school attendance increases with wealth, even in the absence of this conditionality.
The Kingdom of Morocco has successfully instituted a series of reforms in the educational system aimed at achieving universal primary school enrollment. From 1999 to 2006, the enrollment rate increased from 79% to 88% at the national level and from 58% to 87% in rural areas. According to recent surveys, however, dropout rates continue to be significant in primary levels. This is particularly true in the five poorest regions of Morocco where this study takes place, and in regions where the poverty rate is above 30%. In order to address these issues, the Higher Council of Education (CSE) together with the National Ministry of Education (MEN) are planning to launch a nation-wide conditional cash transfer (CCT) program to encourage parents to keep their children in school.
To help the CSE and MEN create an optimally designed program, J-PAL is helping evaluate a two-year pilot in five regions. A total of 320 school sectors are involved in the pilot. They were randomly assigned to three groups:
(1) In 180 school sectors, the government is providing funding to the families conditional on school participation. The cash transfer is made every two months to parents of children attending primary school regularly. The monthly allowance increases with progress through school, starting from US$8 for each child in grade 1 and 2 to US$13 for children in grades 5 and 6. The transfer is maintained as long as the child's attendance rate is at least 80%.
(2) In 80 school sectors, the government is providing funding to all families with children of primary school age, whether or not children attend school. The transfers amounts and frequency are similar to those in the conditional group.
(3) Finally, families from 60 school sectors serve as a comparison group and do not benefit from cash transfers.
In order to compare various ways to monitor that the conditions of the transfers are respected, the 180 school sectors with the conditional program were further randomized into three groups:
(A) In 60 sectors, the monitoring is “light”: the teacher records student attendance on the class register and these records are used to determine eligibility for the transfer.
(B) In 60 sectors, the monitoring is of “medium intensity”: attendance registers kept by teachers are also used to determine eligibility, but teachers are threatened by audits. If a register is audited by a government inspector and a teacher is caught misreporting students’ absences in order to ensure their eligibility, the teacher can be sanctioned.
(C) In 60 sectors, the monitoring is “intensive”: attendance is recorded daily with a digital fingerprint recognition time-clock.
Finally, in order to determine if the effectiveness of the transfers depends on the gender of the beneficiar (the child's mother or father), mothers receive the transfers in half of the school districts sampled for transfers and fathers receive the transfers in the other half. Outcomes of interest include student attendance, dropout rates, test scores and household welfare.
Results and Policy Implications
Treatment and comparison groups were targeted geographically, which removed any ambiguity on eligibility and allowed for large program take-up. Around 97 percent of the study households in either the LCT or CCT communities had at least one child enrolled in the program. Compliance with the random assignment of the gender of the transfer recipient was very high: it was close to 89 percent on average in schools where mothers had been designated as recipients, and around 80 percent in schools where fathers had been.
The Tayssir cash transfers greatly increased school participation under all versions of the program, with the LCT having slightly larger impacts. After two years, the dropout rate among students enrolled in school at the start of the program in LCT schools was about 7.6 percentage points lower than the dropout rate in comparison schools (at 10 percent), a 76 percent decrease. Re-enrollment of those who had dropped out of school before the program almost doubled inLCT schools as compared to comparison schools, and the share of students who never enrolled in school fell by 43 percent. Performance on a basic arithmetic test improved but not significantly.
There was no difference in impacts between transfers issued to fathers and transfers issued to mothers. Making cash transfers conditional did not improve the effectiveness of the program either, but may actually have somewhat reduced it: relative to LCT schools, CCT schools had a slightly higher drop-out rate. Among students who were not enrolled at the start of the program, re-enrollment in CCT schools was lower than re-enrollment in LCT schools, perhaps because conditionality discouraged some households or teachers from enrolling weaker children in the program. The rules of the program were overall poorly understood by local communities. After one year, in the CCT groups, only about half of parents interviewed knew transfers were conditional; in the LCT groups, only about half of parents interviewed knew transfers were unconditional. A weak understanding of what households must do to qualify for the cash transfer may weaken the potential for conditionality to matter.
Adding conditionality or directing the cash transfers to mothers did not increase the program’s impact on student attendance or school enrollment. This is likely because the Tayssir program was framed as an educational support program and perceived as an endorsement of the local schools, since headmasters were responsible for enrolling families. Data from the household surveys provides evidence that Tayssir, in all forms, increased parents’ belief that education was a worthwhile investment.
Overall, the results of the Tayssir experiment suggest that in some contexts unconditional but labeled transfers targeted at poor communities can provide parents with the small nudge necessary to increase attendance. This nudge is relatively inexpensive, due to both small transfers per child and small administrative costs.