How do educational transfers affect child labour supply and expenditures? Evidence from Indonesia of impact and flypaper effects

Southeast Asia

Indonesia

Education

Vouchers / Scholarships / School Subsidies

I De Silva, S Sumarto

2015

De Silva, Indunil, and Sudarno Sumarto. “How Do Educational Transfers Affect Child Labour Supply and Expenditures?: Evidence from Indonesia of Impact and Flypaper Effects.” Oxford Development Studies 43.4 (2015): 483–507.
View publication

Secondary Data: Indonesia Social and Economic Survey (Susenas). https://www.ilo.org/surveyLib/index.php/catalog/2064.

2015

Cash Transfer for the Poor Students programme/Bantuan Siswa Miskin (BSM)

The program started in 2008. The data is exclusively from the Indonesia Social and Economic Survey (Susenas) – 2009

The objective of the transfers is to remove barriers preventing marginalized students from participating in education, assist poor students in gaining appropriate access to education services, to prevent school dropouts, help in meeting the educational needs of at-risk children, as well as to support the Government’s Nine Years Compulsory Education program. The purpose of the study is to test for intra-household flypaper effects (IFE) by examining the impact of an education cash transfer on child-level education expenditures in Indonesia.

Eligibility criteria were not fully clear in the program. The authors state that although the targeting for the BSM program lacked clarity in terms of identifying the poor households, the quota or the number of beneficiaries was largely determined by the availability of funds received by provincial authorities from the Ministry of Education, while the selection of beneficiaries was often left to local education offices or headmasters of the schools. At the national level, there was an understanding that the scholarships needed to prioritize children whose families were in the conditional cash transfer program (PKH program, described below) as these families were very poor.

Disbursement of the transfer was done at the beginning of the academic year (usually in July or August). In 2009, the BSM program covered around 3.6 million students. The program coverage increased to eight million students in 2013 across the country, ranging from primary school to tertiary education levels. In 2009, the unit cost per scholarship ranged from IDR360,000 (US$30) to IDR720,000 (US$60) depending on the level and type of school the student was attending. Poor students enrolled in madrasah (Islamic) schools received a higher amount than those in regular schools.

• Children’s labor supply
• Educational expenditures

Propensity Score Matching

Study design: for all child-level educational expenditure analyses, the authors use children from the ages of 6–18 years. Only children who regularly participate in the labor market as child workers are considered. It is assumed that these children supply labor either to earn a living for themselves or to supplement household incomes. Children engaged in housekeeping activities and who perform household chores—such as cleaning, cooking, or washing—are thus not regarded as child labor in this study. Since all work-related questions were asked only for individuals above the age of 10 years, the sub-sample for child labor supply will be for all children between the ages of 10–and 14 years. The model developed specifies the decision for a child to attend school, supply labor or both is a time allocation decision. the decision of whether a child works or attends school is a joint one as both activities would be competing for the child’s time. The authors use a bivariate Probit model that explicitly takes this interdependency into account and tests the likelihood of children working and going to school, conditional on varying individual and household characteristics.
Sample: the same sample was used for both core and module questionnaires (see below). It consisted of 291,753 households and was designed to be representative at the national, province, and district/city levels. The Susenas 2009-July survey includes an education cost data module containing various expenditure categories for each school-enrolled child.
Survey Instrument: The core questionnaire contains basic information

about household and individual characteristics including health, death, education/literacy, employment, fertility and family planning, housing, and household expenditure. There are three modules of Susenas and each module is added in a 3-year cycle. In 2009, the module’s topic was social life, culture, and education.

Evidence that educational cash transfers and related assistance programs decreased the time spent by children in income-generating activities in Indonesia. Households receiving educational transfers, scholarships, and assistance were also found to spend more at the margin on voluntary educational goods. These results were stronger for children living in poor families.

Child Labour

Households

National

Evidence that educational cash transfers and related assistance programs decreased the time spent by children in income-generating activities in Indonesia. The results were stronger for children living in poor families.

Households receiving educational transfers, scholarships, and assistance were found to spend more at the margin on voluntary educational goods. These results were stronger for children living in poor families.

Single survey (no panel data or cross-section).