Small unconditional grants provided an extra year of education for about 89 dollars in cash transfers and 10 dollars in administrative costs.
Finding ways to improve school attendance among low-income students in developing countries is a perennial challenge. In Turning a Shove Into a Nudge? A "Labeled Cash Transfer" for Education (NBER Working Paper No. 19227), Najy Benhassine, Florencia Devoto, Esther Duflo, Pascaline Dupas, and Victor Pouliquen study the results from the Tayssir pilot, a large government-run randomized controlled trial of the influence of small conditional and unconditional grants on school participation in Morocco for the academic years 2007-2009. They find that small unconditional cash transfers explicitly labeled as providing for educational support produced large gains in school participation. The gains occurred even when parents knew that the grants would continue whether or not children attended school, and regardless of whether fathers or mothers received the money. The average annual transfer per household equaled about 5 percent of annual expenditure. The administration costs of the program were kept low by the fact that targeting was community-based: in poor communities selected to receive grants, everyone in the community was eligible. Households only had to sign up at the local school to start receiving the transfers.
Over the two years of the program, the dropout rate for those who received the unconditional grants fell by 70 percent. The program also increased re-enrollment of those who had already dropped out by 85 percent, and cut the share of never-schooled by 43 percent. Scores on a basic arithmetic test improved. One arm of the study included grants that were conditional on school attendance. While those grants also improved participation, their impacts were somewhat smaller. Fewer students who had dropped out re-enrolled, and there was no improvement in arithmetic scores. The authors estimate that in 2008 dollars, Tayssir's small unconditional grants provided an extra year of education for about 89 dollars in cash transfers and 10 dollars in administrative costs.
The authors suggest that the strong results from the unconditional grants are due in part to an "endorsement effect" that occurs when a large pro-education government program enters poor communities: survey data show that this led parents to update their beliefs about the returns to education. They also point out that by removing ambiguity about who could receive benefits in a given geographic area, the grants program generated a benefit take-up that ranged between 73 and 97 percent.