Evaluating CTs’ impacts

Update on impact evaluations of CT programs

Update on impact evaluations of Sub-Sahara African CT programs

Since the writing of The Cash Dividend, several evaluations have used experimental or non-experimental methods to try to determine the impact of cash transfer programs in Sub-Saharan Africa on key outcomes of interest. Highlights from some of these evaluations are as follows:

South Africa’s Child Support Grant

(Methods: propensity score matching, double differences/differences, dose response)
  • Children who receive the grant in their first two years of life are more likely to have their growth monitored, and those whose mother had over 8 grades of schooling also had higher height-for-age.
  • Enrollment in the CSG at birth results in 10.8 percent (0.38 grades) higher schooling and scores on a mathematics test than enrollment at 6 years old.
  • Girls’ enrollment in the CSG at birth rather than age 6 also reduced school entry delays by 27 percent and increased earlier enrolling girls’ math and reading scores. This result was stronger in households where mothers had less than 8 grades of education.
  • Early enrollment in the CSG also decreased children’s incidence of illness, especially for boys and for children whose mothers had 8 or more grades of education.
  • When anyone in the household received the CSG, adolescents (particularly males) had lower school absences.
  • Individuals who benefitted from the CSG in their first seven years of life were less likely to work outside the home as adolescents.
  • The CSG reduced adolescents’ use of risky behaviors related to alcohol and drug use, crime, involvement in gangs, sexual activity, and pregnancy. Results related to alcohol and drug use, sexual activity, and pregnancy were stronger when children received the grant in early childhood.

Malawi SCT

(Methods: propensity score matching, difference-in-difference)
  • Beneficiaries increased ownership of agricultural tools such as hoes, axes, and sickles, and of livestock, particularly goats and chickens
  • Households reduced participation in agricultural wage labor and ganyu (low-skilled, low pay work)
  • Children reduced labor outside of the household, shifting labor in favor of household chores and activities as adults increased home productive activities.
  • Households reduced negative risk-coping strategies such as begging and taking children out of school to work. They also received fewer gifts from others, a strategy often used to help desperate households cope with risk.

Kenya CT-OVC

(Methods: randomization, difference-in-differences regression)
  • There was a 6.73 percentage points lower probability of sexual debut, particularly for male beneficiaries. The CT also reduced risky sexual behaviors.
  • Beneficiary households increased dietary diversity: they increased spending on meat and dairy, and decreased spending on cassava.
  • The CT changed spending patterns for health, transportation, communication, alcohol, and tobacco, potentially due to the labeling of transfers for the use of children (i.e., a very general soft condition stating the OVC is the responsibility of the parent). More specifically, beneficiaries increased spending on health and decreased spending on alcohol and tobacco.
  • The CT did not impact school enrollment for primary age beneficiaries unless these children faced relatively high access costs. Beneficiaries who lived over 2 kilometers from a primary school increased enrollment by 19 percentage points, and children attending a school with a higher cost of primary school index increased enrollment by 6 percentage points and were 0.31 fewer grades behind.
  • There was a significant 7.8 percentage point increase in secondary school age children’s enrollment. Secondary school age beneficiaries also were fewer grades behind in school and more likely to progress to the next grade. Secondary school age children located farther than 2 kilometers from their school also were 0.324 fewer grades behind due to the CT.

Burkina Faso CCT/UCT

(Methods: randomization, difference-in-differences)
  • Conditional, but not unconditional, cash transfers increased school enrollment and attendance after two, but not one, years of transfers. The insignificant result after one year was attributed to the timing of initial program roll-out.
  • When transfers were given to mothers, both conditional and unconditional transfers significantly increased school enrollment and attendance. The impact was not present in transfers given to fathers.

Niger’s Zap it to Me Program

(randomization, difference, difference-in-differences)
  • Mobile phone distribution decreased costs required for the implementer and beneficiaries to access transfers.
  • Mobile phone transfer households had better outcomes than those who received cash by hand. These households spent cash on a wider array of products, depleted fewer non-durable assets, ate a wider array of foods and cultivated a wider variety of crops. These results are attributed to the lower costs associated with receiving cash through the mobile phone mechanism, the greater privacy mobile transfers provided to beneficiaries, and women’s increased household power.
In addition, some ongoing studies will inform policy makers’ understanding of cash transfers and outcomes particularly important to Sub-Saharan Africa. These include the Swa Koto, South Africa, CCT, which provides cash to a parent/guardian and girls grades 8-11 for girls’ education, with the goal of reducing the incidence of HIV; the Reducing HIV in Adolescents (RHIVA) CCT in South Africa, which transfers cash to males and females in grades 9 and 10 conditional on school attendance, passing exams, improving academic performance, participating in after school programs, and HIV testing, to reduce HIV incidence and decrease risky sexual behaviors; Tanzania’s Iringa Combination HIV Prevention Trial, a UCT given to a parent/guardian and child, which is hoped to decrease HIV incidence and improve sexual behavior; and an evaluation of Zambia’s Vulnerability Grant, a UCT, on sexual behavior of males and females 13-17 years old. Results from these CTs, which are all experimental with the exception of the Zambian program, will provide additional rigorous evidence on how CTs can be used to fight HIV in Sub-Saharan Africa.