Their place is in school and not in the factories. Still, millions of children work, many in hazardous conditions, to support their families.

A low-cost insurance innovation for the very low-income households in Hyderabad, Sindh, has been successful in limiting child labour. The National Rural Support Program (NRSP), Pakistan’s premier poverty alleviation agency, which focuses on rural development, devised innovative insurance regimes for those who borrowed from the NRSP’s microfinance schemes.

The results of a carefully designed experiment, which appeared in the Journal of Health Economics, revealed that households with the enhanced insurance reported lesser incidence of child labour than the rest.*

Pakistan gained notoriety for child labour with the unfortunate death of Iqbal Masih, a teenage boy who was shot dead in 1997. While he died in mysterious circumstances, there was nothing mysterious about him being a child labourer. His death led to a boycott of Pakistani-made carpets, which many believed employed child labour.

Also read: Those underpaid, underage workers of factories

Child labour continues to be ubiquitous in Pakistan. The high fertility rates among the very poor serve as an insurance policy when the breadwinners die suddenly or are incapacitated because of illness or injury. The children are shipped to work to support the family when the traditional support structure breaks down.

Banning child labour has been the preferred unthinking solution for many. However, such drastic interventions have seldom delivered. In fact, such bans condemn children to even worse outcomes, such as prostitution.

The NRSP came up with a novel solution that targeted the root cause of the problem, and not merely its symptoms. Knowing that children are sent to work when the support network collapses, the NRSP intervention bolstered the safety net against unexpected hazards.

The results of the intervention are interesting. Households with enhanced insurance options reported less incidence of child labour and lower school dropout rates than the rest.

NRSP's insurance experiment

Starting in 2005, the NRSP mandated insurance for those who borrowed from their microcredit schemes. The insured included the borrower, the spouse, and children under 18 years of age.

In order to offer improved insurance services, the NRSP in 2009 conducted an experiment in Hyderabad, Sindh. In nine branches randomly selected out of 13 branch offices, the NRSP offered optional insurance (in addition to the mandatory insurance) for other adult members of the family (that also included children over 18 years), for a nominal sum of 100 rupees. In addition, NRSP visited borrowers each month to determine if they needed help with filing insurance claims. The four remaining branch offices offered only the mandatory insurance without the two enhancements.

Also see: For child labourers, education still a distant dream

In total, almost 2,097 households (13,000 individuals) were part of the experiment, of which 1,320 households were in the ‘treatment’ group comprising the nine branch offices that received enhanced insurance options. Another 777 households were in the control group which was offered only the basic insurance.

The analysis revealed that despite the fact that the treated and control communities were comparable before the start of the experiment, the communities recipient of the enhanced insurance options reported significantly less incidence of hazardous occupations for child labourers than the rest did. Furthermore, child labour earnings were also lower for communities with enhanced insurance options than the rest.

A breakdown by gender presented even more pronounced results for boys, who are more likely than girls to be pushed out to work. In addition to lower hazardous occupation and child labour earnings, communities with enhanced insurance options reported higher school attendance for boys. The same was not true for girls.

Read on: Satyarthi — a campaigner against child slavery

The study in the Journal of Health Economics though, suffers from a few shortcomings. The primary concern is about the aggregation bias. Instead of estimating the empirical models at the household level, the authors aggregated the household data at the community level and did empirics on the aggregated data comprising 13 observations. The other limitation is that the authors focused on the communities, thus merging the results for households that bought the enhanced insurance and those who did not in the nine treated communities.

The result:

Still, the results obtained are based on sound reasoning. Child labour occurs when the social safety net is missing for low income households, who need support when the primary breadwinner could no longer provide.

The NRSP’s insurance scheme offers the buffer, especially when the interruption is temporary, to households that need help.

Given the tremendous burden of disease carried by the low-income households in Pakistan, there is merit in devising a similar national insurance scheme for the very poor.

*Landmanna, Andreas and Markus Frölicha, Can health-insurance help prevent child labor? An impact evaluation from Pakistan. Journal of Health Economics. 39 (2015) 51–59.