Skip to main content

Evaluation of the social cash transfer pilot programme, Tigray region, Ethiopia-endline report (Berhane et al. 2015)

Absence of conflict of interest.

Citation

Berhane, G., Devereux, S., Hoddinott, J., Hoel, J., Roelen, K., Abay, K. Kimmel, M., Ledlie, N., & Woldu, T. (2015). Evaluation of the social cash transfer pilot programme, Tigray region, Ethiopia-endline report. New York, NY: UNICEF

Highlights

  • The study’s objective was to examine the impact of a conditional cash transfer program on child labor and school enrollment in two districts in Ethiopia (Abi Adi and Hintalo Wajirat).
  • The study used a nonexperimental design to compare the outcomes of children who received the cash transfer to those who did not. Outcomes were measured using baseline and follow-up household surveys and analyzed using regression.
  • The study found that the number of days that girls were engaged in household business (non-farm activities) significantly decreased in Abi Adi; no significant impacts were found for boys in Abi Adi or any children in the Hintalo Wajirat. The study also found that the rates of school enrollment significantly increased for girls aged 6-11 in Hintalo Wajirat; there were no significant impacts for boys in Hintalo Wajirat or any children in Abi Adi.
  • The quality of causal evidence presented in this report is moderate because it was based on a well-implemented nonexperimental design. This means we are somewhat confident that the estimated effects are attributable to the conditional cash transfer program and not to other factors.

Features of the Intervention

The Social Cash Transfer Pilot Program (SCTPP) was introduced in 2011 to alleviate poverty and improve the quality of life for vulnerable children, older persons, and persons with disabilities. Households were selected for the program if they were extremely poor and labor constrained (i.e., elderly head of household with no working adults aged 19-60, one adult is responsible for three dependent members, or contains a chronically ill or disabled adult). Eligibility was initially determined by Community Care Coalitions (CCCs) and neighborhood (tabia) authorities. Two CCC members visited the households and ranked them from most to least needy. A community meeting was held to review the list. Social workers from the district (woreda) then made a final visit to verify eligibility. Households received monthly payments of 25 birr for each child under 16 years of age plus an additional 10 birr if the child is enrolled in school; payments were made for up to four children. An additional payment was provided if the household contained a disabled child under 18 years of age (40 birr), a disabled adult (50 birr), or an elderly dependent (60 birr). The minimum payment amount each month was 155 birr (approximately $8 USD).

Features of the Study

The study was a nonexperimental design conducted in two districts, the town of Abi Adi and a rural area, Hintalo Wajirat. In Abi Adi, all neighborhoods were included. In Hintalo Wajirat, 7 of the 22 neighborhoods were initially included. Another neighborhood, Bahr Tseba, was added eight months after the program started when additional funding was secured. Neighborhoods were purposefully selected based on proximity to the highway to make program implementation easier and reduce administrative costs. Within Abi Adi, all smaller regions were included. In Hintalo Wajirat, three were excluded due to geography (e.g., rugged terrain, lack of roads, etc.).

The authors created three groups: 1) a treatment group selected from households that received SCTPP benefits, 2) a control group comprised of households that were eligible for SCTPP benefits, but were not enrolled primarily due to lack of funds, and 3) a random sample of households that did not meet eligibility criteria. The primary reason for the random sample was so that the authors could determine how accurately eligible households were targeted, and to identify any larger trends in the neighborhoods or districts included in the study. However, the random sample was not used in impact analyses. There were 1,696 households in the treatment group, 1,539 households in the comparison, and 432 households in the randomly selected group. The authors oversampled non-elderly households in the treatment group to ensure that a sufficient number of children were included in the sample.

The first household survey was conducted between May and June of 2012 and included 3,667 households. The follow-up household survey was conducted between May and July of 2014 and included 3,351 households. Child labor outcomes included the number of days spent in wage labor and household business labor (nonfarm) and hours spent on household chores at the child and household level; school enrollment was measured as enrollment at any time in the school year. Impacts were assessed using regression and disaggregated by district (Hintalo Wajirat or Abi Adi), age (6-11 or 12-16), and gender.

Findings

Working children/Child labor

  • The study found that the number of days that girls were engaged in work in household businesses (non-farm activities) significantly decreased in Abi Adi by one day. There was no significant impact on boys in Abi Adi, or on boys or girls in Hintalo Wajirat.
  • The study did not find significant impacts on wage labor for boys or girls in Abi Adi or Hintalo Wajirat.
  • The study also did not find significant impacts on household chores in Abi Adi or Hintalo Wajirat.

Education (School participation/enrollment)

  • The study found that school enrollment of girls aged 6-11 in Hintalo Wajirat significantly increased by 13.3 percentage points. There was no significant impact on older girls (aged 12-16) or boys in Hintalo Wajirat, or on boys or girls in Abi Adi.

Considerations for Interpreting the Findings

The first survey occurred eight months after the SCTPP started, so it was not a true baseline survey. During the first household survey, the authors included several recall questions for key outcomes that asked respondents to go back 12 months to May 2011 before the program/payments began. They then used single difference models for data analysis when appropriate.

Causal Evidence Rating

The quality of causal evidence presented in this report is moderate because it was based on a well-implemented nonexperimental design. This means we are somewhat confident that the estimated effects are attributable to the Social Cash Transfer Pilot Program and not to other factors.

Reviewed by CLEAR

February 2021

Topic Area