Absence of conflict of interest.
de Hoop, J., Kovrova, I., & Rosati, F. C. (2016). Eliminating child labour in El Salvador through economic empowerment and social inclusion: Impact report. Retrieved from http://www.ucw-project.org/attachment/07032017224El_Salvador_IE_07042016_web.pdf
- The study’s objective was to assess the effects of a women’s entrepreneurship training program on children’s participation in school and work.
- The study was a regression discontinuity design, using baseline and follow-up household surveys to measure program outcomes. Households were selected into the program if they had a wealth score below a cut-off point. The authors compared the outcomes of children in households above and below the cut-off score to understand the program’s effects.
- The study found that selection and participation in the program was significantly associated with a decrease in the percent of children working-only and not attending school, and the number of hours that children worked per week. Selection and participation in the program was also significantly related to an increase in school attendance.
- This study used a regression discontinuity design and therefore was reviewed using CLEAR’s descriptive study evidence review guidelines. As such, it does not receive a causal rating.
Eliminating Child Labour in El Salvador through Economic Empowerment and Social Inclusion
Features of the Intervention
The Eliminating Child Labour in El Salvador through Economic Empowerment and Social Inclusion program was implemented by the International Labour Organization, World Bank, and UNICEF to reduce child labor. The program took place in five El Salvador municipalities: Tacuba, Izalco, San Luis la Herradura, Tecoluca, Santiago Nonualco. These five municipalities were selected because they were located relatively close to one another, and they met criteria of having a high rate of child labor, being sufficiently safe for field workers, and having enough residents to conduct the study. The program provided entrepreneurship training and a business starter kit for mothers. The women were also offered training to help transition students who were not attending school back to school, but this component was not utilized by many participants. Households were eligible to participate if at least one of the children in their household was working and aged 5-17, and also if there was at least one woman in the household (aged 18-65) who was literate but did not have salaried employment.
Features of the Study
The study used a regression discontinuity design (RDD), where participating households were selected into the treatment group based on whether they were below a cut-off score on a wealth index. The authors used baseline survey data to create the wealth index; however, they did not specify the value of the cut-off point that was used. Those with a lower score than the cut-off were eligible for the treatment group that received program services (1,098 households) and those with a higher score were in the comparison group that did not receive any program services (1,000 households). The study sample included children who were aged 5-15 at the time of the baseline survey. A follow-up survey was administered 3 years after the baseline survey to assess child work and school participation outcomes. For the follow-up survey, the sample included 794 treatment households and 702 comparison households. In addition to assessing any work participation for children aged 5-13, the study assessed work for more than 34 hours per week or in hazardous conditions for children aged 14-15, or work in hazardous conditions for children aged 16-17. The authors used a RDD statistical model to compare the outcomes of the groups above and below the cut-off score for wealth to assess the impact of being selected for the program. The outcome analyses were focused on the households within a limited range (or bandwidth) around the wealth cutoff score. The study presented intent-to-treat analyses (which can be interpreted as the impact of being offered the program) in addition to analyses of those who participated in the program (treatment-on-the-treated).
- The study found that selection and participation in the program was not significantly related to children’s rates of employment or rates of working in hazardous conditions.
- For children’s allocation of time between school and work, the study found that selection and participation in the program was significantly associated with lower rates of working-only and not attending school. However, selection and participation in the program was not significantly related to rates of studying-only or rates of both working and studying.
- The study found that selection and participation in the program was significantly related to a decrease in the number of hours that children worked per week.
Education (School participation/enrollment)
- The study found that selection and participation in the program was significantly associated with higher rates of school attendance. However, selection and participation in the program was not significantly related to the regularity of children’s school attendance.
Causal Evidence Rating
This study used a regression discontinuity design and therefore was reviewed using CLEAR’s descriptive study evidence review guidelines. As such, it does not receive a causal rating.