Absence of conflict of interest.
Citation
Del Carpio, X. V., Loayza, N. V., & Wada T. (2016). The impact of conditional cash transfers on the amount and type of child labor. World Development, 80, 33-47. doi:10.1016/j.worlddev.2015.11.013
Highlights
- The study’s objective was to examine the impact of a conditional cash transfer (CCT) program on the amount and type of child labor conducted by children ages 8 to15. Types of child labor included household chores, farm labor (both at home and outside the home), and skill-forming activities defined as work in commerce, sales, or manufacturing activities outside the home.
- The study was a randomized controlled trial conducted in six municipalities in Nicaragua. Households from each community were randomly assigned into treatment and control groups. Treatment households received either the basic CCT or the basic CCT plus an additional a household grant for the creation of a micro business. Survey data were collected and child labor outcomes were analyzed using two statistical models.
- The study found that the program significantly reduced overall child labor, household chore labor, and farm labor for both treatment groups compared to the control group. The program had no effect on skill-forming labor for the treatment group that received the basic CCT but increased skill-forming labor for the children in the households that received the basic CCT plus the business grant.
- The quality of causal evidence presented in this report is high because it was based on a well-implemented randomized controlled trial with low attrition. This means we are confident that the estimated effects are attributable to the Atención a Crisis program and not to other factors.
Intervention Examined
Atención a Crisis (“Attention to the Crisis”)
Features of the Intervention
Atención a Crisis was a conditional cash transfer program implemented between November 2005 and December 2006 in six municipalities affected by a severe drought and classified as extremely poor in northwest Nicaragua. The program aimed to help poor households deal with the negative effects of the drought, as well as improve household economic outcomes. Households would receive payments, conditional on school attendance and health check-ups for children in the household. The households receiving the intervention were to receive one of three packages: a basic cash transfer; the basic cash option plus a scholarship for one member of the household 16 or older to attend occupational training; or the basic cash option plus a household grant to create a non-agricultural micro-business. The value of the cash transfers was between 25 percent and 45 percent of pre-program annual household income. Due to budget constraints, only 3,000 households could be treated under the program. Ultimately the scholarship option was not provided.
Features of the Study
The program was designed as a randomized controlled trial to examine the effects of cash transfers on the amount and type of child labor conducted by children (ages 8-15). All communities from six municipalities were paired based on socio-economic and geographic factors and randomized into treatment and control communities based on a lottery. Within the treatment communities, the poorest 90 percent of households were offered the program. The invited households were then assigned by lottery to one of the treatment conditions. Ninety-five percent of invited households participated, but in order to avoid selection bias, the analysis was conducted on all invited households regardless of participation One thousand households were randomly selected from the control communities to represent the control group.
Originally, 3,000 treatment households were assigned to one of the three intervention packages (cash transfer, cash plus scholarship, or cash plus business grant). The scholarships for the second group ultimately were not provided, so the first and second conditions were combined. The baseline sample included 2,323 households with children between the ages of 7 and 14 (1,136 in the basic treatment group, 580 in the basic plus business grant treatment group, and 607 in the control group). Baseline survey data were collected in May/June 2005 and the follow-up occurred in August/September 2006 (hence the age range of 8-15 for analysis). The data collection months correspond to the same agricultural season and similar school period in Nicaragua. The authors used two statistical models to evaluate the effect on labor for children age 8-15.
Findings
Employment/Child labor
- The study found a significant decrease in child labor hours among those in households receiving the CCT and the business grant compared to the control group. Children in these households worked, respectively, 1.8 and 0.9 fewer hours per week than those in the control group.
- Children from households receiving the basic CCT worked less than those who received the CCT plus the business grant. Children in both the basic CCT and the CCT with the business grant worked fewer hours in household chores and farm labor than the control group, but children in the CCT with the business grant worked more in skill-forming labor than children in the control group or the basic CCT.
- Children in the group receiving the CCT and business grant worked about one (1) less hour per week in household chores and 0.5 fewer hours in farming than children in the control group. Children in the basic CCT group worked about 0.9 fewer hours per week in household chores and 0.7 fewer hours in farming than children in the control group.
- For children in the basic CCT group there was no statistical difference in the number of hours spent on skill-forming labor compared to those in the control group, but children from households also receiving the business grant worked about one (1) hour more per week in skill-forming activities than those in the control group (and those receiving the basic CCT).
- All findings were statistically significant unless otherwise noted.
Considerations for Interpreting the Findings
The study does not report attrition separately for each group. However, the overall household attrition between baseline and follow-up was 1.3 percent and it is unlikely that the difference in the rate of attrition between the treatment and control groups would exceed 5.8 percent.
Causal Evidence Rating
The quality of causal evidence presented in this report is high because it was based on a well-implemented randomized controlled trial with low attrition. This means we are confident that the estimated effects are attributable to the Atención a Crisis program and not to other factors.