Absence of conflict of interest.
Citation
de Hoop, J., Ranzani, M., Rosati, F. C. (2014). The impact of the Kenya CT-OVC programme on children’s activities (Working Paper). Retrieved from http://www.ucw-project.org/attachment/The_impact_of_the_Kenya_CT-OVC20140423_144552.pdf
Highlights
- The purpose of the study was to assess the impact of the Kenya Cash Transfer for Orphans and Vulnerable Children Program on child labor and school participation.
- The study used a randomized controlled trial to assess the intervention. Child outcomes were compared between the treatment and control group households, using data from follow-up surveys administered two years after the start of the intervention.
- The study found a 5 percent reduction in the proportion of children in the treatment group who were working in unpaid labor or conducting household chores compared to the control group.
- The quality of causal evidence presented in this report is low because randomization was compromised and the authors did not ensure that the groups being compared were similar before the intervention. This means that if there were significant effects, we would not be confident that they were attributable to Kenya CT-OVC; other factors are likely to have contributed.
Features of the Intervention
The Kenya Cash Transfer for Orphans and Vulnerable Children Program (Kenya CT-OVC) provided cash transfers to households with vulnerable children, many who had at least one deceased parent. Households participating in the program received a monthly cash transfer valued at $24 USD. Initially, the plan was that some households would participate in a Conditional Cash Transfer (CCT) version of the program, where certain requirements needed to be met to receive the transfers, or would participate in an Unconditional Cash Transfer (UCT) version of the program, where those requirements did not need to be met. In the CCT version, the requirements would include regular school attendance, visits to health clinics, and adult educational sessions. However, across communities, there were varying levels of restriction on whether households needed to meet certain conditions; these variations were not measured. As a result, the authors combined the CCT and UCT households into one treatment group. Eligible households met the following criteria: were poor, were not participants in another cash transfer program, and had one or more OVC. An OVC was a child younger than 18, who fit into at least one of the following categories: had a deceased parent, had a chronic illness, lived in a household headed by a child, or had a caregiver with a chronic illness.
Features of the Study
The study used a randomized controlled trial and took place in 7 districts in Kenya: Nairobi, Kwale, Garissa, Homa Baye, Migori, Kisumu, and Suba. Within each of the seven districts, four localities were selected. Two of these localities were randomly assigned to the treatment group and two were randomly assigned to the control group. Within the treatment communities, eligible households were selected into the study through a multi-step process: first, a committee administered a survey at households to assess whether they were eligible. Then, households that appeared to be eligible were asked to complete a longer survey to affirm that they were, in fact, eligible. While eligible households were asked to attend an event to enroll, the authors noted that in some communities there were more eligible household than there were program resources; in those cases, they prioritized households with the youngest and oldest caregivers. Households enrolled in the treatment group received the monthly cash transfer, and those in the control group did not. This paper focused on 4,843 children ages 4-15 at the start of the study (2,068 households), whose household had a survey at both baseline and follow-up. The authors compared outcomes of whether children participated in paid work and whether they attended school between the treatment and control groups and controlled for baseline group differences such as age, highest level of education among household adults, and ownership of land and livestock. The participation in paid work and school attendance outcomes were measured using data from a two-year follow-up survey.
Findings
Working children/Child labor
- Participation in Kenya CT-OVC was significantly associated with a 5 percent reduction in the proportion of children who participated in unpaid work or household chores. Using a composite measure of any work (paid or unpaid) or household chores, program participation was also significantly associated with a 5 percent reduction in the proportion of children compared to the control group.
- There was no statistically significant relationship between participation in the Kenya CT-OVC and the hours in unpaid work or household chores compared to the control group.
- There was no statistically significant relationship between participation in the Kenya CT-OVC and the proportion of children who participated in paid work, as measured using data from the two-year follow-up survey.
Education (School participation/enrollment)
- There was not a statistically significant relationship between participation in the Kenya CT-OVC and the proportion of children who attended school, as measured using data from the two-year follow-up survey.
Considerations for Interpreting the Findings
Random assignment was compromised for this study because, after randomly assigning the localities to the treatment or control group, the authors used different procedures in each study group to identify households to participate in the study. In treatment localities, two rounds of screening questionnaires were used to confirm eligibility, and then households were prioritized giving favor to those with the youngest and oldest caregivers. Families then had to decide to attend an enrollment event. Control households were selected based on their baseline characteristics, but not using the same procedure as the treatment households. This implies that the households in the control group likely differed from those in the treatment group, defeating the intent of random assignment. In addition, the authors did not account for preexisting differences between the groups before program participation, such as child labor and child gender. These preexisting differences between the groups—and not the Kenya CT-OVC —could explain the observed differences in outcomes.
Causal Evidence Rating
The quality of causal evidence presented in this report is low because randomization was compromised and the authors did not ensure that the groups being compared were similar before the intervention. This means that if there were significant effects, we would not be confident that they were attributable to Kenya CT-OVC; other factors are likely to have contributed.