Absence of conflict of interest.
Hoddinott, J., Gilligan, D. O., & Taffesse, A. S. (2009). The impact of Ethiopia’s Productive Safety Net Program on schooling and child labor. Retrieved from http://ssrn.com/abstract=1412291
- The purpose of the study was to assess the impact of Ethiopia’s Productive Safety Net Program (PSNP) on child labor and school attendance, along with the potential added benefit of participation in the Other Food Security Program (OFSP). This summary focuses on the comparison between the group receiving the PSNP benefit and the comparison group.
- The study used a nonexperimental design to compare the outcomes of children ages 6-16 that received the cash transfers with those who did not, based on data from the Food Security Program Survey. Using several demographic characteristics, the authors created a matched comparison group of households who did not receive the benefit to assess the effectiveness of the cash transfer program.
- The study did not find a significant relationship between receipt of benefits from PSNP and the number of hours worked (in domestic labor, agricultural labor, or total labor hours) or school attendance rates.
- The quality of causal evidence presented in this report is low because the authors did not ensure that the groups being compared were similar before the intervention.This means we are not confident that the estimated effects are attributable to the Productive Safety Net Program; other factors are likely to have contributed.
Features of the Intervention
Ethiopia’s Productive Safety Net Program (PSNP) was implemented to improve food security, and maintain household and community financial assets. The program operated in 262 districts in four regions of Ethiopia: Amhara, Oromia, Tigray, and the Southern Nations, Nationalities and Peoples Region (SNNPR). In the program, household members could earn money (6 birr or $.61 USD) for each day they work on community development projects, such as building roads. Payments were typically paid monthly, but in some areas they were only given once the community development project was complete. The Other Food Security Program (OFSP) was also offered in some localities at the same time as the PSNP. The OFSP offered additional benefits to support food security such as advice on farming crops and access to credit. Communities largely selected households to participate in the PSNP based on whether they were poor and had received food aid in the past, although the specific selection process varied by locale.
Features of the Study
The study used a matched comparison group design to understand the effects of the PSNP program on school attendance and child labor for children ages 6-16. Since the OFSP program was expected to also help improve food security, the authors were also interested in assessing the combined effect of participating in both the PSNP and OFSP program. About 18 months after the PSNP began, the Food Security Program Survey was administered. The survey asked households about participation in public safety net programs such as the PSNP and OFSP, child and household characteristics, and children’s schooling and labor outcomes. The survey included questions about the households at the current time of the survey, and also included retrospective questions about the households prior to the initiation of the PSNP. Using propensity score matching, the authors created comparison groups matched on pre-program characteristics such as age, gender, asset levels, and schooling. The study included three treatment groups and comparison groupsbased on survey responses about household participation in the programs:
- Treatment group 1 included households who had received any payment from the PSNP program within the past 12 months. A matched comparison group was developed for that treatment group that included any households that had not participated in PSNP or had participated but not received any payment through that program.
- Treatment group 2 included households that received at least 90 birr per household member (or about $9 USD) through the PSNP. The matched comparison group developed for that group included households who did not participate in the PSNP program.
- Treatment group 3 included households that received any payment from PSNP and received any type of benefit from OFSP. The matched comparison group included households that did not receive any OFSP benefits, and that was either not a participant of PSNP or was a PSNP participant who received no payments from the program.
All treatment and comparison groups excluded households who received Direct Support, which was a program that provided transfers to the elderly or people with disabilities. The authors compared school attendance and child labor at 18 months between three treatment groups of those participating in the PSNP and those in a corresponding matched comparison groups who did not participate in the PSNP. Child labor was measured on the survey by asking households for the number of hours that children worked in domestic chores, agricultural labor, and in those two categories plus paid labor outside the home. School attendance was measured by asking households whether or not children attended school at the time the survey was completed. Analyses were separated to determine program effects by gender and child age (ages 6-10, 11-16, and 6-16).
Working children/Child labor
- For boys and girls of all ages, the study did not find a significant relationship between receipt of benefits from PSNP and the number of hours worked in domestic labor, agricultural labor or in total labor including paid labor outside the home.
Education (School participation/enrollment)
- For boys and girls of all ages, the study did not find a significant relationship between receipt of benefits from PSNP and school attendance rates.
Considerations for Interpreting the Findings
While the study used a matched comparison group design, the authors did not account for other factors that could have affected the difference between the treatment and comparison groups, such as prior child labor participation. Also, the matched comparison group was formed using data collected 18 months after the intervention through retrospective survey items; households may not have reported accurately on their baseline characteristics given how much time had passed since the program began.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because the authors did not ensure that the groups being compared were similar before the intervention. This means we are not confident that the estimated effects are attributable to the Productive Safety Net Program; other factors are likely to have contributed.