Child labour persists not because parents do not value education, but because the economic calculus at the household level often makes children's labour more immediately valuable than their schooling.
In a tribal household with three children, minimal landholding, and seasonal income that peaks at harvest, a child who works earns ₹100–200 per day in agricultural labour at the peak season — money that may be the difference between whether the family eats during the lean months. The child who attends school produces no immediate income and requires expenditure (transport, materials, uniform, informal fees). The opportunity cost of schooling is real and large.
Adding to this, the quality of the school — particularly in remote tribal areas — is often poor enough that parents have rational grounds for doubting whether the investment will pay off. A school where the teacher is absent half the time, where no child can read by Class 5, and where the curriculum is delivered in a language the children don't speak is not obviously a better use of the child's time than agricultural labour that produces immediate, certain income.
UNICEF Innocenti's 2024 analysis of child labour and schooling in India is explicit on this: "Overall trends in child labour have declined as education enrolment has increased — suggesting a strong association between universal education and eradication of child labour — many forms of child work and child labour continue to persist, just as school attendance and completion rates reveal gaps in educational attainment." The direction is right: better education reduces child labour. But better enrolment without better quality is not sufficient.
This economic framing matters for programme design. An approach that removes children from labour without addressing the household economic pressure that put them there will see children return to work as soon as the programme's direct intervention ends. Rescue is not a programme strategy. It is an emergency response. The programme strategy is addressing the economic conditions that make child labour rational.
Why Rescue-Only Fails
The National Child Labour Project (NCLP), India's primary child labour intervention, has been rescuing and mainstreaming children for over three decades. In 2021–22 and 2022–23, the scheme rescued and rehabilitated 18,137 and 13,761 children respectively — a genuine achievement.
And yet child labour persists. The US Department of Labor's 2023 India findings note explicitly that "many children have re-entered the labor market as local schools cannot accommodate all released children and do not provide the services that former child labourers need." Rescue followed by mainstreaming into a school that doesn't work for the child produces re-entry into child labour.
The structural problem with rescue-only approaches:
They address the symptom, not the cause. A family whose income depends on a child's labour will find the next child — or the same child — returning to work as soon as the rescue pressure lifts. The family's economic situation has not changed.
They treat children as passive victims. Children who have been working, often for years, have agency and identity connected to their work. They have peer networks. They have skills. Removing them from work without addressing their social and psychological context produces trauma, not rehabilitation.
They rely on enforcement in settings where enforcement is structurally inadequate. Labour inspectors are spread thin. Child labour is seasonal and mobile. Agricultural child labour in remote tribal areas is essentially invisible to formal enforcement systems. A programme that depends on enforcement to produce outcomes in contexts where enforcement cannot reach will produce outcomes only in the contexts it can reach — which are typically not the highest-concentration areas.
The Dual-Track Model: What the Evidence Supports
The evidence from UNICEF Innocenti's review of educational strategies for reducing child labour, and from the broader international child labour literature, points toward interventions operating at four levels simultaneously. This is the dual-track approach — addressing both the demand side (making children's labour less economically necessary) and the supply side (making schooling more valuable and accessible) at the same time.
Track 1: The family and household level
Conditional and unconditional cash transfers: The evidence on cash transfers for reducing child labour is positive and consistent. Families that receive conditional transfers (conditional on school attendance) or unconditional transfers (provided regardless of behaviour) reduce child labour and increase school attendance. India's own evidence on this — from JSY (which incentivises institutional delivery), MGNREGS (which reduces the marginal value of child agricultural labour by providing adult employment), and various scholarship schemes — supports this relationship.
MGNREGS as a child labour reduction tool: A consistently underappreciated finding from India's research: areas with higher MGNREGS implementation have measurably lower child labour prevalence, because MGNREGS provides adult wage employment that substitutes for child labour in household income strategies. For NGOs in tribal Odisha, ensuring that households in high-child-labour areas are accessing MGNREGS entitlements is a legitimate child labour intervention. Mapping entitlements (covered in the Entitlements Mapping Practice Note) and MGNREGS access are directly relevant here.
Family economic support: Scholarships for children attending school; links to livelihood programmes for parents; social protection scheme access — all reduce the household economic pressure that makes child labour necessary. The child's value as an economic contributor falls when the household has alternative income sources.
Track 2: The child level
Scholarships and incentives: Direct incentives for school attendance — scholarships, bicycle grants, uniforms — reduce the opportunity cost calculation that keeps children out of school. For children already in work, bridge programmes that allow them to catch up on learning while transitioning back to school are more effective than forcing immediate re-enrolment in age-grade classrooms where they are hopelessly behind.
Life skills and legal rights awareness: Children who understand that they have a right to education, and that their labour may be illegal, are more likely to assert that right — particularly adolescents who have enough agency to influence family decisions. This is not about encouraging confrontation with parents; it is about giving children the knowledge that enables them to participate in household decisions about their own future.
Bridge schooling for withdrawn children: NCLP residential schools — designed specifically for children withdrawn from child labour — provide bridge education that prepares children for mainstreaming into government schools. Where they function well, these schools are a critical component of the withdrawal-to-rehabilitation pathway. Where they function poorly — as many do — mainstreaming into a functioning government school with targeted academic support is often more effective.
Track 3: The school level
School quality improvement: The evidence from UNICEF Innocenti is direct: school quality is a mediating factor between child labour and education. Schools that children and parents value — where they learn, where the teacher is present, where the learning environment is welcoming — draw children away from work. Schools that don't function keep children in work by failing to provide a compelling alternative.
For NGOs: improving school quality in high-child-labour communities is a child labour intervention. TaRL implementation (Practice Note 16), SMC activation (Practice Note: SMC Activation), and teacher attendance monitoring are all relevant.
Flexible scheduling: Where agricultural child labour is seasonal — concentrated in harvest and sowing seasons — schools that provide catch-up programmes and flexible scheduling for children who miss school during peak agricultural demand retain more children over time than those that treat any absence as a dropout. This requires advocacy with block education officials, but is achievable in communities with active, functioning SMCs.
Track 4: The community and system level
VLCPC activation: The VLCPC (covered in the VLCPC Activation Practice Note) is the mandated community-level child protection institution. An active VLCPC that identifies child labour cases, tracks children who have dropped out of school, and connects families to economic support schemes is the community-level prevention mechanism.
Community norm change: In some communities, child work is normalised — it is what children do, it is understood as skill transmission, it is not perceived as harmful. Norm change is slow and requires sustained community engagement over multiple years. It is also necessary, because enforcement alone cannot change community norms.
PENCIL Portal: The government's PENCIL portal (pencil.gov.in) is the coordination mechanism for child labour enforcement — allowing complaint filing, case tracking, and programme monitoring. For NGOs, understanding how to use the PENCIL portal to file complaints about child labour is a practical skill that complements community-level prevention work.
Odisha-Specific Considerations
Agricultural child labour in tribal districts: Most child labour in tribal Odisha is agricultural and seasonal — children involved in paddy transplanting, tobacco cultivation, and sugarcane harvesting, often alongside their parents. This labour is partly invisible (happening within the family or on community land) and partly seasonal (concentrated in specific months). Programmes designed for urban or industrial child labour — factory rescues, urban domestic service — do not fit this context.
Migration-linked child labour: Seasonal migration from Odisha's tribal districts for construction, brick kilns, and cane harvesting in other states takes both adults and children. Children who migrate with parents are out of school for the migration period — three to six months in some cases. Some migrate alone or with contractors, which is trafficking regardless of whether it appears voluntary. VLCPCs with migration-season protocols (discussed in the VLCPC Practice Note) are the primary prevention mechanism.
Bonded child labour: While rare compared to agricultural and domestic service child labour, bonded labour among tribal children — typically in brick kilns and stone quarries, often linked to debt bondage — requires specific legal intervention. The Central Sector Scheme for Rehabilitation of Bonded Labourers provides for release certificates and rehabilitation support. NGOs encountering bonded labour situations should contact the District Labour Office and the Sub-Divisional Magistrate, who have authority to issue release certificates.
What NGOs Should Focus On
For NGOs without dedicated child labour mandates, the most effective interventions are the complementary ones: ensuring families access economic support schemes (Entitlements Mapping), activating VLCPCs, improving school quality in high-labour-concentration communities, and monitoring for migration-linked child labour in seasonal out-migration periods.
For NGOs with dedicated child labour mandates, the dual-track integrated model is the evidence-based approach: child-level incentives and bridge schooling combined with household economic support, school quality improvement, and community-level VLCPC activation. Any single component in isolation will underperform.
The measure is not how many children are rescued. It is how many children are still in school, and not working, twelve months after the initial intervention.
Related Knowledge Commons content: Child Welfare Sector Primer (Sector 01) · Practice Note: VLCPC Activation · Practice Note: Entitlements Mapping · Practice Note: Teaching at the Right Level
Evidence Grade: B — Multi-study. This Practice Note draws on UNICEF Innocenti's Child Labour and Schooling in India (2024), UNICEF Innocenti's Educational Strategies analysis (2024), US Department of Labor India 2023 and 2024 reports, and CRY India's integrated child labour programme documentation. Last reviewed: April 2026.
Questions or corrections: knowledge@jabasu.org