School Attendance, Child labour, and Cash Transfers

Latin America and the Caribbean

Uruguay

Social Protection

Conditional Cash Transfers (CCT)

Amarante, V., M. Ferrando, and A. Vigorito

2011

Amarante, V., M. Ferrando, and A. Vigorito. “School Attendance, Child Labor, and Cash Transfers. An Impact Evaluation of PANES.” PIERI Working Paper, 2011.
View publication

-

2011

PANES

April 2005 – December 2007

“To provide direct assistance to households that had experienced a rapid deterioration in living standards since the 2001-2002 crisis” and to “strengthen the human and social capital of the poor”.

All households with per capita income below approximately US$50 per month (computed as the maximum of social security income and self-declared income) could apply for participation in the program. Among these households, eligibility was determined on the basis of a predicted poverty score, which was calculated using baseline household socioeconomic characteristics. An separate eligibility threshold was determined for each of the country’s five main administrative regions (the paper does not explain how this threshold was established).

PANES included several components. The largest was a monthly cash transfer of US$56 per month per household. This transfer amount equals approximately 50% of average self-reported pre-program household income. Households with children or pregnant women were also entitled to a food card, and in-kind transfers (the paper does not provide details on the monthly value of these transfers). Other smaller components of the program included a workfare program, job training, adult educational interventions, and health care subsidies (the paper does not provide further explanation on these components). PANES was conditional on children’s school attendance and health check-ups. However, due to limited coordination between institutions, conditionality was not enforced. In fact, 42% of beneficiary households was not aware that participation in the program was even nominally conditional.

• Education: School attendance of 3 to 17 year-old children (no definition given)
• Children’s Work: Child labor of 6 to 17 year old children (no definition given)

Regression discontinuity design

Identification Strategy: The paper exploits the threshold scores in the predicted poverty scores in a sharp regression discontinuity framework.
Data: The analysis is based on administrative baseline data collected to determine eligibility and a follow-up survey which was administered between March and June 2008 (shortly after the program ended).
Sample: 3000 households with predicted baseline poverty scores near the eligibility threshold.

Education: Regression discontinuity estimates suggest that the program had no effect on school attendance. This result is robust to different specifications and persists when children are disaggregated by age or gender. Children’s Work: Regression discontinuity estimates suggest that the program had no effect on child labor. This result is robust to different specifications and persists when children are disaggregated by age or gender.

Child Labour

Households

National

The program seems to have had no effect on child labour.

The program seems to have had no effect on school attendance.

The authors show that there is no robust evidence that awareness of the school participation condition resulted in a significant effect of the program on school attendance.