Absence of conflict of interest.
- The study’s objective was to examine the impact of the Productive Social Safety Net (PSSN) program on child labor and school participation outcomes. This summary focuses on the comparison of the conditional cash transfers with Public Works Program component treatment group versus the control group.
- The study was a randomized controlled trial where villages were randomly assigned to one of three study arms: conditional cash transfer (CCT) only, CCT with a supplemental public works program (PWP), and the control condition. The authors used difference-in-differences models to compare the changes in outcomes between the groups.
- The study found that the probability of participation in paid work outside the household significantly decreased for children in PWP households compared to children in control group households. However, the probability of participation in livestock herding significantly increased for children in PSSN households compared to children in control households. Also, the likelihood of dropping out of school significantly decreased for children in PSSN households compared to children in control households.
- The quality of causal evidence presented in this report is high because it was based on a well-implemented randomized controlled trial. This means we are confident that the estimated impacts are attributable to the Productive Social Safety Net program, and not to other factors.
Features of the Intervention
The Productive Social Safety Net (PSSN) program was the flagship social protection program of Tanzania. The PSSN was initiated in 2012 by the Government of the United Republic of Tanzania and implemented by the Tanzania Social Action Fund (TASAF). The PSSN aimed to reduce child labor and improve youth well-being and transitions to adulthood by providing cash transfers to poor and vulnerable households. The PSSN provided bimonthly payments to households with both an unconditional base transfer as well as additional amounts conditional on health check-ups and children's school attendance. The program also contained a Public Works Program (PWP) component that gave households an opportunity to earn additional income. The PWP component of the program is a 'cash-for-work' service to help supplement household incomes during the lean season. The PWP provides other work opportunities as well as a fixed per day grant for one able-bodied adult per household aged 18 and over for up to 60 days in a four-month period.
Features of the Study
The study was a randomized controlled trial that took place across 102 villages from the mainland of Tanzania and Zanzibar. Eligible villages and households within those villages had to meet strict poverty guidelines in order to be considered for inclusion in the study. A three-stage targeting process was used where the poorest villages were identified; the poorest households within those villages were identified; and the households were subjected to a proxy means test to ensure they meet the poverty criterion. The authors randomized the 102 villages into three different study arms. Thirty-five villages received conditional cash transfers (CCT) only, 26 villages received CCTs and the supplemental public works program (PWP), and 41 villages received the control condition (delayed participation in PSSN by 18 months). Within each village, eligible households were randomly selected using a list of targeted households for the TASAF program.
The baseline study sample consisted of 4,134 children aged 4-16 from 801 households across the 102 villages. The endline sample consisted of 3,360 children aged 5-17 (1,937 children in the treatment group and 1,423 in the control group). Using data from household questionnaires, the authors conducted difference-in-differences models to compare the outcomes of treatment and control members between baseline in 2015 to endline in 2017. Only individuals who were interviewed at both baseline and endline were included in the analyses. Although there were two different treatment groups, the authors tested overall impacts between each treatment group and the control group as well as comparing outcomes between the pooled PSSN treatment (CCT only and CCT + PWP) and control groups. Child labor was measured as: (1) any economic activity in the past year, (2) paid work outside the household, (3) work in the household business (non-agricultural), (4) farm work for the household, (5) livestock herding for the household, and (6) fishing for the household. School participation was measured as currently attending school, attended school regularly in the past week, and dropped out of school between baseline and endline.
Working children/Child labor
- The study found that the probability of participation in paid work outside the household significantly decreased by two percentage points for children in PWP households compared to children in control households.
- However, the study found that the probability of participation in livestock herding significantly increased by four percentage points for children in PWP households compared to children in control households.
- The study did not find significant differences between children in the PWP group and the control group in any economic activity in the past year, work in the household business, farm work for the household, or fishing for the household.
Education (School participation/enrollment)
- The study did not find significant differences between children in the PWP group and the control group in their rates of attendance at school.
- However, the study found that children in PWP households were significantly less likely to drop out of school between the baseline and endline surveys compared to children in the control group.
Considerations for Interpreting the Findings
The authors had a well-designed experimental study, but anticipation effects are something to consider when interpreting the findings. The control group was a delayed treatment group where participants received a conditional cash transfer 18 months later. If participants in the control group knew these benefits were forthcoming, it is possible their awareness impacted their behavior and thus had an undue effect on the results.
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. This means we are confident that the estimated impacts are attributable to the Productive Social Safety Net program, and not to other factors.