Intergenerational Conflict and Schooling Decisions in Brazil

Researchers:
Lucas Coffman
Location:
Brasilia, Brazil
Sample:
210 families
Timeline:
2009 - 2009
Target group:
  • Children
  • Parents
  • Students
Outcome of interest:
  • Enrollment and attendance
Intervention type:
  • Conditional cash transfers
  • Monetary incentives
AEA RCT registration number:
AEARCTR-0001596

Even though many developing countries have adopted policies to increase school attendance, many primary school-aged children in developing countries do not attend school. One explanation for low attendance is that parents and children may disagree about the benefits of schooling and students may choose to skip classes when their parents are not monitoring their attendance. A study on parents’ demand for conditionality in a Brazilian cash transfer program revealed that parents are willing to pay to for the conditionality in order to monitor their children’s school attendance.

Policy issue

In an effort to achieve the U.N. Millennium Development Goal of universal primary school enrollment, many developing countries have adopted policies to increase school attendance. However, the United Nations reports that 11 percent of primary school-aged children in the developing world do not attend school.1 One explanation for low attendance is that parents and children may disagree about the benefits of schooling, as parents often value schooling more than their children. In this case, students may choose to skip classes when their parents are not monitoring their attendance.

One method of monitoring and increasing primary school attendance is through conditional cash transfers (CCTs). CCT programs provide low-income households with money as long as they meet certain requirements, such as sending their children to school. This allows the programs to simultaneously address poverty and income inequality, and incentivize schooling by encouraging parents to better monitor their children’s attendance.

Context of the evaluation

This evaluation took place in low-income favelas in the Brazilian capital of Brasilia. These neighborhoods are often characterized by high levels of criminal activity, poverty, and a lack of essential public services. Despite compulsory education laws in Brazil, over 9 percent of 14-year-old children from the lowest-income households were not enrolled in school in 2006.2 In 2008, 10 percent of the population aged 15 years and older was non-literate, and the average number of years of schooling for individuals aged 10 and older was 7.1.3 In response, the Brazilian government implemented a CCT program known as Bolsa-Escola Vida Melhor (School Stipend, Better Life). At the time of this study, Bolsa-Escola provided each beneficiary household with R$120 per month (approximately US$74) for each child between the ages of six and 15, as long as the child attended a minimum of 85 percent of school days that month. This represents a significant amount of income for beneficiary households, whose monthly income averaged below R$195 per month (approximately US$120).

Details of the intervention

Researchers designed the evaluation to indicate the extent to which parents are willing to pay for Conditional versus Unconditional Cash Transfers as a means to monitor their children’s behavior, and to reveal any intra-household conflicts regarding school attendance. They conducted a randomized evaluation during June and July 2009, in which 210 Bolsa-Escola beneficiary families were randomly selected to participate in a survey. Only families with children aged 13 to 15 who had no older CCT-eligible children were permitted to participate. Parents were offered R$7 or R$10 to take the survey. Parents and children participated in the study. Before being surveyed, families were randomly divided into four groups:

Treatment group Intervention description
Baseline The parent was asked to choose between his or her current CCT and an Unconditional Cash Transfer (UCT) of increasing size, until he or she made the switch. The parent was then asked the same questions with the UCT at the base value and the CCT increasing. The child would be informed of any change in conditionality.
Don't tell The parents were offered the same choices as in the baseline. However, their child would not be informed about any program change, and thus would not know whether the conditionality had been removed.
Text message The parents were offered the same choices as in the baseline. They were also offered the option of receiving free cell-phone text messages each time the child missed school regardless of the parents’ choice between conditional and unconditional payments. This gave parents complete information of their child’s attendance pattern regardless of their choice for the cash transfer.
Non-classroom The parents were offered the same choices between CCTs and UCTs. However, the conditionality would now be on whether the child was present on school grounds during the day, with no obligation to attend classes. This was implemented to potentially observe how parents value the non-classroom content of school, such as keeping the child off the streets.

Parents were also informed that 5 percent of study participants would be randomly selected to have one of their decisions implemented.

Results and policy lessons

Baseline treatment: In this group, 82 percent of parents chose a CCT over a UCT of greater value, and were willing to forego at least R$37 per month to keep conditionality. Children were significantly less likely to be willing to pay for conditionality. These results suggest that parents place a high value on their child’s school attendance, and believe that the conditions placed on the transfers can help encourage school attendance.

Don’t tell treatment: When children were unaware that the conditionality on their attendance had been removed, parents were significantly less likely to demand that conditionality. The portion of parents willing to pay for conditionality decreased to 30 percent from 82 percent in the Baseline group. This suggests that parents value the ability to monitor their children’s attendance and believe their children do not value education as highly as they do.

Text message treatment: When parents were alerted of their child’s attendance status by a text message, and thereby provided monitoring information, they were also significantly less likely to demand conditionality. The proportion of parents who were willing to pay for conditionality decreased to 34 percent from 82 percent in the Baseline group.

Non-classroom treatment: Finally, when conditionality is changed from mandating classroom attendance to mandating presence on school grounds only, parents did not significantly decrease their demand for conditionality relative to the Baseline group. The proportion of parents who were willing to pay for conditionality decreased slightly. These results indicate that schooling may have benefits beyond providing in-classroom learning opportunities, such as keeping children off the streets.

The results from all four treatment groups indicate that parents and children disagree about schooling decisions and that, in this context, parents are willing to pay to monitor their children’s school attendance. This study suggests that households need efficient mechanisms to close the information gap between parent and child, such as with text messaging, and provides an additional rationale for adopting CCTs for the monitoring they provide. By better informing children of the value of education, policymakers may be able to bridge parent-child differences in schooling preferences.

Bursztyn, Leonardo and Lucas Coffman. 2012. "The Schooling Decision: Family Preferences, Intergenerational Conflict, and Moral Hazard in the Brazilian Favelas." Journal of Political Economy 120(3): 359-397.

1.
United Nations Millennium Development Goal Report (2011). United Nations. New York, NY.
2.
PNAD (2006) Pesquisa Nacional por Amostra de Domicilios, Institute Brasileiro de Geografia e Estatisica. Rio de Janiero, RJ.
3.
PNAD (2008) Pesquisa Nacional por Amostra de Domicilios, Institute Brasileiro de Geografia e Estatisica. Rio de Janiero, RJ.