Incentivising school attendance in Mozambique: Reports or cash transfers?

An enormous effort has been made by governments the world over to incentivise parents to ensure that their children attend school regularly. In developing countries, in particular, arguably the most significant innovation in social policy in the past few decades has been the introduction of conditional cash transfers made to parents to incentivise a number of prescribed behaviours such as regular school attendance.

There is evidence that cash transfers can have a positive impact on school attendance. However, incentivising children using report cards of their school attendance may be a more cost-effective approach. When children have a say in decisions regarding attendance (e.g., because their parents do not have full information as to whether the child is present at school), simply providing information to the parent about their child’s attendance at school may improve attendance independently of any cash transfer.

The study

Damien de Walque and Christine Valente provide experimental evidence in a recent study from Mozambique. It reveals that children have agency in decisions regarding their schooling as early as sixth grade. Neglecting this reality may lead to inefficiencies in education policies.

Cash transfers are becoming more common in Africa, but evidence is still scarce on the following questions that are the focus of a recent pilot co-funded by the World Bank and International Growth Centre (IGC):

• Can results be achieved with a cheaper intervention in which parents simply receive information on their child’s attendance, but no financial incentive?
• How much more effective is a traditional conditional cash transfer at incentivising parents compared to simply providing information?
• How much more effective than simply providing information is a voucher system incentivizing children?

Education in Mozambique

Despite large increases in enrolment rates in lower primary school grades, most children are still not completing primary education in numerous African countries. In Mozambique, only about 10% of rural children and 20% of urban children who start school finish lower primary school (EP1) at the intended age of ten years old. For upper primary schooling (EP2), the official completion rate is abysmal. In rural areas, by age 19 only about 14% of males and 8% of females have completed EP2.

Despite recent progress, this issue is even more pronounced for girls than boys, which is problematic not only from an equity point of view, but also from an efficiency point of view.  Positive externalities from the schooling of women are believed to be larger than those emanating from the schooling of males due to the former’s traditional role as the main caregivers in the household.

Experimental set-up

The experiment focused on Manica province, where excess female drop-out is above the national average. The research team split 173 “complete” primary schools (i.e., schools offering all seven grades of primary education) randomly between four experimental study arms (one control and three treatment groups) using a random number generator.

1. Report cards:

• In each of the three treatment groups, the team provided schools with attendance “report cards” for each girl in senior primary grades (6^th and 7^th Grade, or EP2).
• These simple report cards had a coding easily understood by parents, even if they were illiterate: the teacher drew a circle for a given day if the girl attended school that day, or the teacher marked a cross for each day missed.
• The report cards were given to the girls at the end of each week to show to their parents (who had been informed of the report card system ahead of the intervention). They were then brought back to school at the start of the next week.
• Where conditional transfers were introduced (“Parent Cash” and “Girl Voucher” study arms), a 90% attendance rate condition The verification of this condition relied on the attendance rate calculated based on the report card information.

1. The schools were randomly distributed within the following four experimental study arms:
2. “Girl Voucher”: Vouchers were given to girls in the two final years of primary school (Grades 6 and 7), who could then use the vouchers to buy a selected number of education related items such as: a school uniform, shoes, school bag, smaller materials (pens, notebooks, etc.), which were delivered to the school by the research team. The nominal value of the transfer was 400 Meticais per trimester or USD8.4 per trimester at the start of the experiment. Therefore, the maximum annual transfer was worth 1200 Meticais or eight times the daily wages of an agricultural labourer in the study area.
3. “Parent Cash”: The research team gave money to the parents and made the same items available as in the “Girl Voucher” treatment arm for optional purchase at the school. The value of the transfer was the same as in “Girl Voucher” treatment arm (400 Meticais per trimester).
4. "Report Card": Girls simply received the weekly attendance report card.
5. Control Group.

Findings

• The main result of this experimental study is that the information content (report card) of a conditional transfer can have a substantial effect on school attendance independent of any cash transfer.
• The estimated effect of the “Report Card” treatment on attendance is as large as 54% of the “Girl Voucher” incentive effect and 75% of the effect of the “Parent Cash” incentive.
• The effect of the “Parent Cash” incentive on attendance is not, from a statistical point of view, significantly different from that of the “Report Card” only.
• Incentivising children with the “Girl Voucher” is significantly more effective than simply providing information with the “Report Card” (and nearly twice as effective).
• Rewarding children is at least as effective as rewarding their parents.
• Transfers to children were not appropriated by their parents.
• The study replicated findings from most evaluations of conditional cash transfers (to parents) that gains in attendance do not translate into gains in test scores.
• However, both the “Report Card” and the “Girl Voucher” treatment incentives improved scores in a math test.

Policy implications

1. Given the low cost of the “Report Card” treatment and the ease with which the treatment could be scaled, this is a promising policy option for the many governments lacking the administrative and budgetary capacity to implement a conditional cash transfer programme.
2. Children have agency in the decision to attend school which suggests that policies taking this into account could be more effective than traditional policies focusing on parents as the sole decision makers.
3. Although the nominal value of the parent and children transfers was the same, incentivising children is also potentially more cost-effective than incentivising parents (provided transport costs are not excessive) because they were incentivised with vouchers to be redeemed against a limited number of items that can be procured in large volumes at a discounted price.
4. It is concerning that gains in attendance do not translate into gains in test scores because schooling is only a means to an end – that of acquiring cognitive and other skills.
5. Improved parental information and attendance are both beneficial for cognitive skills, but conditional cash transfers directed at parents may have counterproductive effects on learning. 

References:

de Walque, D. and Valente, C. (2018). Incentivizing School Attendance in the Presence of Parent-Child Information Frictions, Policy Research Working Paper 8476, Development Economics Development Research Group, World Bank Group. Accessible: http://documents.worldbank.org/curated/en/720071529003623702/pdf/WPS8476.pdf