Poverty alleviation and child labour

South America

Ecuador

Social Protection

Unconditional Cash Transfers (UCT)

EV Edmonds, N Schady

2012

Edmonds, Eric V, and Norbert Schady. “Poverty Alleviation and Child Labor.” American economic journal. Economic policy 4.4 (2012): 100–124.
View publication

Primary Data: N. A. Data collected by the Catholic University of Ecuador.

2012

Bono de Desarrollo Humano (BDH)

June 2003-March 2005 (baseline and endline surveys)

The program aims to support the human capital of poor children. The study’s objective is to examine the impact of cash transfers on children’s outcomes (age range: 11-16 years old).

The poorest two-fifths of the population were eligible for the BDH. Starting in 2001, the government developed a family means test, called the Selben index. The Selben index was computed by first conducting a census of household assets, then using principal component analysis to create a single asset index.

The evaluation of the BDH randomly assigned eligibility for the cash transfer to poor eligible households. The BDH transfer was $15 per household per month, 7 percent of monthly expenditures for recipient households. The amount did not vary across eligible families. The transfer was paid to mothers and did not have any conditions attached. The rollout of BDH explicitly contained a randomized component in four of Ecuador’s 24 provinces. Within provinces selected for the evaluation, parishes were randomly drawn. A parish is a unit of local government with an average population of 26,503. Within selected parishes, a sample of 1,488 households was randomly selected for the evaluation. There were three filters for incorporation into the evaluation sample. First, households had to be BDH-eligible. Second, households that already received Bono Solidario transfers, and continued to be eligible for the BDH program, were excluded from the evaluation—for these households, there was no policy change, and therefore nothing to evaluate. Third, households had to have at least one child, age 6–17, at the time the means test data were collected. Households in the evaluation sample were randomly assigned to a “treatment” group (lottery winners) and a “control” group (lottery losers) with a random number sorter. On average, there were 14 lottery winners aged 11–16 at baseline and 13 lottery losers per evaluation parish. Lottery winners would immediately be activated for transfers.

• Child labor (hours of work, paid and unpaid)
• School enrollment

RCT

Study design: The BDH lottery is independent of employment opportunities and the household’s time allocation decision-making process. Hence, the lottery solves the problem of confounding factors and simultaneity that plague most of the literature on the impact of poverty on child labor. The empirical strategy was straightforward. The randomization was stratified by parish, so the authors included parish fixed effects throughout and clustered robust standard errors by parish. With the inclusion of parish fixed effects, the empirical approach only captures effects of the BDH that are net of any spillovers to the control population.
Sample: Within selected parishes, a sample of 1,488 households was randomly selected for the evaluation.
Survey Instrument: The main sources of data used in this paper are the baseline and follow-up surveys designed for the BDH evaluation. Both surveys were carried out by the Catholic University of Ecuador, an organization that had no association with the BDH program and no responsibility for its implementation or evaluation. The survey instrument includes a roster of household members, basic socio-demographic characteristics of these members, detailed time allocation and schooling information for school-age children, employment status for adults, dwelling characteristics, household asset holdings, and an extensive module on household expenditures, following the structure of the 1998–1999 Encuesta de Condiciones de Vida (ECV).

Poor families with children in school at the time of the small cash transfer (less than 20% of median child labour earnings) use the transfer to postpone the child’s entry into the labor force. Students in families induced to take up the transfer by the experiment reduced paid employment by 78 percent and unpaid economic activity inside their home by 32 percent. Time in unpaid household services increases, but overall time spent working declines.

Child Labour

Households

National

An unconditional cash transfer program in Ecuador targeted at poor women with children led to the postponement the child’s entry into the labour force. Students in families induced to take up the transfer by the experiment reduced paid employment and household chores. Time in unpaid household services increased, but overall time spent working declined.

-

The paper finds a larger magnitude of the BDH effects on child time allocation relative to other estimates in the literature. The reason for this is that a much larger fraction of the children in the sample work in paid employment than is the case in other studies, creating a larger margin for declines in Ecuador than elsewhere. Also, the estimated treatment effects apply only to those who won the lottery (self-selection is not possible).