Crop Choice, School Participation, and Child labour in Developing Countries: Cotton Expansion in Burkina Faso

Africa

Burkina Faso

Labour Market

Agricultural productivity

H Kazianga; F Makamu

2017

Kazianga, Harounan, and Francis Makamu. “Crop Choice, School Participation, and Child Labor in Developing Countries: Cotton Expansion in Burkina Faso.” American Journal of Agricultural Economics 99.1 (2017): 34–54.
View publication

Secondary data is available at: https://microdata.worldbank.org/index.php/catalog/2538/get-microdata.

2017

N.A.

1985-2005 (range of data observed)

Policy reform in 1995 to expand cotton farming, which happened for certain regions and not for others.

Child labor is examined indirectly.

The authors use the cotton expansion (done through a cotton policy reform in 1995) to identify the effects of cotton farming on school participation and child labor in the new cotton region. At baseline, about 21.8% of households reported farming cotton in the old cotton region, and 2.4% of households reported farming cotton in the new cotton region. In the noncotton region, about 3.1% of households farmed cotton. After the policy reform (1998–2003), the percentage of households farming cotton increased to 40.3% in the old cotton region and 11.3% in the new cotton region. The percentage of households that reported farming cotton decreased slightly in the noncotton region.

• Income effect: increase in income for regions that had an expansion of cotton farming compared to the region that was not subject to it.
• Child participation in the labor market.
• Children’s schooling (enrollment rates, number of years in school).

Difference-in-Differences

Study design/Identification Strategy: The cotton expansion to the new region was decided at the central government level and was essentially given to local households. In particular, there is no evidence that local households and decision-makers initiated the policies that induced the central government to expand cotton production in their areas.
Sample: the authors use data from three rounds of the national priority surveys conducted in 1994, 1998, and 2003. The three surveys are similar in the scope of the information collected, sampling design, and coverage. The surveys are closely related to the World Bank Living Standards Measurement Study (LSMS).
Survey Instrument: All three surveys use a two-stage stratified random sampling. The analysis focuses on rural households while the subsample of urban households is used for robustness checks. The surveys collected information on household and individual characteristics, employment status, expenditures, and income. Information on school enrollment status and participation in various activities was collected for school-age children.

Households that were exposed to the policy reform had higher consumption expenditures, while those in the old cotton area did not experience any significant change. There have been substantial gains for girls in the new cotton region relative to girls in the noncotton region. Enrollment rates for girls increased by 28.5% and years of education completed increased by 30% (according to the main specification of the paper). Girls’ participation in farm labor decreased by 56.3%. The policy change did not have a statistically significant effect on boys’ participation on farms.

Child Labour

Households

Local

Girls’ participation in farm labour decreased by 56.3% in response to a cotton expansion in Burkina Faso. The policy change did not change boys’ participation on farms.

Households that were exposed to the policy reform had higher consumption expenditures, while those in the old cotton area did not experience any significant change. There have been substantial gains for girls in the new cotton region relative to girls in the noncotton region. Enrollment rates for girls increased by 28.5% and years of education completed increased by 30%.

Several robustness checks were run. Regressions analysis with province fixed effects.