Labour intensive public works and children’s activities: the case of Malawi

Southeastern Africa

Malawi

Labour Market

Public Works

J de Hoop, FC Rosati

2016

de Hoop, J., & Rosati, F. C. (2016). Labor-intensive public works and children’s activities: The case of Malawi. Working Paper.
View publication

N. A.

2016

Malawi’s LIPW program (Labour intensive public works)

2012-2013 (beginning and end of intervention)

Study the impact of a public work program on education and child labor in a randomized controlled experiment.

A "Poverty and vulnerability criteria" and a participatory process involving local officials: the so-called traditional authorities and group village headmen. Within these villages, the group village headmen and local committees jointly allocate the program to households deemed especially vulnerable through an informal process.

The program provides a lump-sum payment in return for temporary work on local infrastructural projects. The projects primarily consist of the construction and "upgrading" of roads, afforestation, and irrigation. A special feature of the program is that it primarily operates during the maize planting season. Sampled eligible households in the treatment groups were invited to participate in two waves of the LIPW program, which varied across the four treatment groups both in terms of the timing of the second wave of the program (lean season versus post-harvest season) and payment schedule (multiple tranches versus lump-sum payment).

• School attendance
• Total expenditures on education
• Children\'s participation in economic activities in the 7 days before the interview and household chores on the day before the interview.

RCT

Study design: the LIPW program offered to the selected households differed randomly across the four treatment groups both in terms of the timing of the second wave of the program and in terms of the schedule according to which beneficiaries were paid. The first wave of the program took place in the 2012 maize-planting season (November and December). The timing of this first wave of the program was identical in all four groups of treatment villages. The second wave of the program took place either towards the end of the 2013 lean season (in March and April) or after the 2013 maize-harvesting period had ended (in June and July). These two groups are referred to as the lean season and post-harvest season treatment villages. Depending on the treatment group to which the village had been assigned, beneficiaries were paid either in one installment approximately a week after completing the work or in 5 installments of 720 Malawi Kwacha each at three-day intervals.
Sample: Villages located in a stratified sample of 12 districts (out of Malawi\'s 28 districts) were randomly selected to be incorporated into the experiment if (i) they had been pre-screened as eligible to benefit from the 2012/2013 wave of the LIPW program and (ii) they had been covered by the World Bank\'s 2010/2011 Integrated Household Survey (also known as the IHS-3) – a nationally representative survey covering 16 randomly selected households in each of 768 villages (or more precisely census enumeration areas). A total of 183 villages were identified as meeting these criteria and incorporated into the sample.
Survey Instrument: Four waves of survey data were collected at relatively short intervals (a few months apart) to capture the differential effects of the four treatment arms.

The LIPW program encourages school attendance of children from households that were offered the program. This effect is particularly pronounced during the lean season when children are especially likely to drop out of school. The finding that the effect on school participation is strongest during the lean season suggests that households use the LIPW program as a buffer during periods of limited resources. Increased school attendance during the planting and (post-) harvest season is also observed. The latter suggests that household participation in the program does not increase the shadow value of children’s productive activities. The program does not have a significant effect on children’s participation in household chores or economic activities, with a couple of exceptions that are not entirely consistent with the estimated effects on education and cannot be easily explained.

Child Labour

Households

National

The program does not have a significant effect on children’s participation in household chores or economic activities in Malawi.

The program encourages school attendance of children from households that were offered the program. This effect is particularly pronounced during the lean season when children are especially likely to drop out of school. The finding that the effect on school participation is strongest during the lean season suggests that households use the LIPW program as a buffer during periods of limited resources. Increased school attendance during the planting and (post-) harvest season is also observed. The latter suggests that household participation in the program does not increase the shadow value of children\'s productive activities.

One limitation in the study is related to the baseline survey. Two sources were used: the IHS-3 (Integrated Household Survey) data, but there was no data available for 24 of the 183 villages. A second baseline was collected in October and November of 2012, but the collection of the second wave of baseline data took place after the LIPW program had been announced in some of the treatment villages and some households (about 6% according to self-reported information in the second baseline) had already started to participate in the first wave of the program.