"Every child in Odisha's tribal belt who grows up malnourished, who leaves school before Class 8, who is trafficked to a brick kiln in Telangana — each of these is not a failure of intent. It is a failure of infrastructure. Of systems. Of the connective tissue between what government promises and what communities receive."
There is a village in Malkangiri district — one of hundreds like it — that the nearest functional primary health centre cannot reach in under three hours by road. The Anganwadi worker assigned to it hasn't been paid in six months. The Child Welfare Committee for the block meets rarely, and when it does, no one from the village knows it exists. The children here are not uncared for — their parents love them fiercely — but they are invisible to every system that was built to protect them.
This is not unusual. This is the norm across Odisha's southern and western tribal districts, home to some of the most acute child deprivation in India, in a state that has simultaneously achieved genuine development milestones. The contradiction is not accidental. It is structural. And understanding why is the first step to changing it.
This Primer draws on global evidence about what actually works in child welfare — not what sounds good in proposals, but what rigorous research across comparable contexts has demonstrated makes a measurable difference in children's lives. It then runs that evidence through Odisha's specific filter: its tribal geography, its government scheme landscape, its cultural realities, and the particular conditions that make some interventions easy to adapt and others nearly impossible to transplant.
The goal is practical: to give NGO leaders, CSR managers, and government programme officers a clear-eyed view of the evidence, so that the next programme in Koraput or Nuapada or Nabarangpur is built on what we know works, not on what we've always done.
Part One: Odisha's Children — The Numbers Beneath the Progress
Odisha has a complicated relationship with its own child welfare story. The state government has, in recent years, made genuine progress on several fronts. Odisha's response to cyclones and disasters — often involving rapid child protection mobilisation — is now internationally regarded. Gross enrolment ratios at primary level have improved substantially. The Mamata scheme, providing conditional cash support to mothers, has been cited as a model maternal health intervention.
But progress at the aggregate level consistently obscures catastrophe at the margins.
As of late 2024, 11,710 children across Odisha were identified as suffering from Severe Acute Malnutrition. The burden falls almost entirely on tribal-dominated districts: Mayurbhanj leads with 1,460 SAM cases, followed by Keonjhar, Nabarangpur, Rayagada, Kalahandi, and Malkangiri. These are not random statistics. They are the districts where 13 Particularly Vulnerable Tribal Groups (PVTGs) live, where forest department restrictions limit traditional food access, where the distance to a functional health facility is measured not in kilometres but in hours.
Among Odisha's PVTG children under five, ICMR research found the prevalence of underweight at 75.26%, stunting at 55.42%, and wasting at 60% — all forms of undernutrition higher among girls than boys. Read that again slowly. Three-quarters of PVTG children under five are underweight. More than half are stunted. These figures place Odisha's most marginalised tribal children in the same statistical territory as the world's most acute humanitarian crises — not as a footnote to India's development story, but as a central fact about what the state's growth has failed to reach.
UNICEF's assessment of Odisha's children captures the core dynamic: tribal children bear a higher burden of stunting and severe wasting, food insecurity is most acute among the most disadvantaged tribal groups, and children's vulnerability deepens with high levels of poverty, rigid social norms, displacement, frequent natural calamities, and the presence of Left-wing extremism in some areas.
Beyond nutrition, the picture is equally layered. Child labour persists — concentrated in brick kilns, carpet weaving, construction, and domestic service, with Odisha consistently among the top source states for inter-state trafficking of children and adolescent girls. Child marriage rates in tribal districts remain well above national averages. Out-of-school children, particularly among Scheduled Tribe communities in blocks with extreme geographic isolation, remain a stubborn reality even as the state's headline enrolment numbers improve.
A fact-finding report from Keonjhar and Jajpur districts captured the lived reality: nearly 90% of rural tribal households live in extreme distress with no liquid cash, communities have little information about their entitlements, and community-based organisations are limited or non-existent in most tribal villages.
What the data does not capture is what happens between the scheme on paper and the child in the village. The Anganwadi system exists; its workers are often underpaid, undertrained, and overwhelmed. The Child Welfare Committee framework exists; its members often have no idea what their authority is. The government schemes — ICDS, Mamata, the National Food Security Act, the Sponsored Scheme for Children — exist with budgets and beneficiary targets. Research examining access to nutrition schemes in Odisha found that limited awareness, lack of proactive disclosure of scheme information, excessive distance from centres, caste-based power dynamics, and weak monitoring institutions were the key factors restricting access of specific social groups to these programmes.
The gap between entitlement and receipt is where most of Odisha's child welfare NGOs do their actual work. Understanding that gap — and what it takes to close it — is what this Primer is about.
Part Two: What the Global Evidence Actually Says
Child welfare is one of the most studied fields in international development. Decades of research across South Asia, Sub-Saharan Africa, and Latin America have produced a reasonably clear picture of what works. Some of it is intuitive. Some of it is genuinely counterintuitive. And some of what "feels right" to well-meaning programme designers turns out to have little or no evidence behind it.
The First 1,000 Days Are Not a Metaphor
The period from conception to a child's second birthday — approximately 1,000 days — is the most biologically sensitive window in human development. The brain grows faster in this period than at any other point in life. Nutrition deficiencies during this window cause structural changes to the brain and body that are, to a significant degree, irreversible. Stunting acquired in these 1,000 days is not simply a height problem — it is associated with reduced cognitive capacity, lower lifetime earnings, higher vulnerability to chronic disease, and in turn, higher rates of malnutrition in the next generation.
This is the most robust finding in all of child development research, and it has a deeply uncomfortable implication: programmes that intervene after age two are operating on a body and brain that has already been shaped by what came before. Supplementary feeding for a four-year-old matters. It is not the same as adequate nutrition for a pregnant woman and a child in the first two years.
The evidence-based response to this is now well-established globally. It involves: quality antenatal care with adequate iron and folic acid supplementation; skilled birth attendance; exclusive breastfeeding for the first six months; appropriate complementary feeding from six months; micronutrient supplementation; and treatment of acute malnutrition when it occurs. The challenge is not the protocol — it is the system required to deliver it consistently to the last mile.
What this means for Odisha: In Malkangiri's most remote blocks, the woman who will give birth in four months has likely received fewer than the recommended antenatal check-ups. The Anganwadi that should be supplementing her nutrition and that of her toddler may be functioning inconsistently. The community health worker who is supposed to support her breastfeeding may be managing a caseload that makes regular home visits impossible. Understanding the first 1,000 days is not enough — what Odisha needs is the infrastructure that consistently delivers within that window, even in the hardest-to-reach places.
Community-Based Management of Acute Malnutrition: The Evidence Gap in India
Globally, community-based management of acute malnutrition (CMAM) has transformed the treatment of severe malnutrition. The core innovation — treating uncomplicated severe acute malnutrition with ready-to-use therapeutic food (RUTF) at community level rather than hospitalisation — was validated in clinical trials across Africa and has since been adopted in WHO protocols. It reduces mortality from severe acute malnutrition dramatically, while being deliverable at community level without hospital infrastructure.
India has been notably slow to adopt CMAM at scale, preferring to route SAM treatment through Nutrition Rehabilitation Centres (NRCs). This is not entirely wrong — NRCs provide intensive treatment — but NRCs are inaccessible for many PVTG families who live hours from the nearest facility. The ICMR research on Odisha's PVTGs explicitly notes that context-specific problems that prevent effective implementation of programmes should be investigated, and that implementation research is needed to improve outcomes from existing government programmes.
What this means for Odisha: NGOs working on child nutrition in tribal blocks have an important role in piloting community-based approaches — not to replace the NRC system, but to extend the reach of nutrition rehabilitation into communities that will never access a hospital. This requires working with community volunteers, Anganwadi workers, and ASHAs in ways that most government schemes currently do not support.
Cash Transfers Work — But the Devil Is in the Design
Conditional cash transfer programmes — where families receive income support tied to specific behaviours like school attendance or health check-ups — have produced some of the most rigorous evidence in development economics. Brazil's Bolsa Família programme, Mexico's Progresa (later Oportunidades), and numerous programmes across Sub-Saharan Africa have demonstrated consistent positive effects on school enrolment, health service utilisation, and nutritional outcomes for children.
More recently, evidence has emerged for unconditional cash transfers. A 2024 study of Flint, Michigan's Rx Kids programme — which provided unconditional cash transfers to every expectant mother in the city — found significant reductions in child welfare system involvement in the first six months of life, suggesting that economic security during the perinatal period directly reduces child malnutrition and neglect. While the Flint context is vastly different from a tribal block in Odisha, the underlying mechanism is universal: child malnutrition is substantially a function of household poverty, and addressing household poverty directly is a legitimate and evidence-backed child welfare intervention.
India's Mamata scheme — providing ₹5,000 to mothers in two instalments linked to antenatal and postnatal behaviours — operates on the conditional cash transfer model. Research in Odisha has found that access to Mamata is restricted by limited awareness, weak monitoring, and caste-based exclusion, meaning the most vulnerable families are often the least likely to benefit.
What this means for Odisha: NGOs have a critical facilitating role in ensuring that cash transfer schemes reach who they are designed to reach. This means community-level awareness campaigns in mother-tongue languages; support for enrolment and documentation; and monitoring that flags exclusion early enough for it to be corrected. It also means building the evidence base for whether these transfers are actually producing the behavioural changes they are designed to produce — the data on Mamata's outcomes, at the district level, is limited.
Child Protection Systems: Why Top-Down Approaches Consistently Fail
The global evidence on child protection systems — the formal architecture of Child Welfare Committees, Juvenile Justice Boards, and child protection units — is both encouraging and sobering. These structures matter. When they function, they provide children with access to legal protection, social services, and rehabilitation support that is otherwise unavailable.
They very frequently do not function.
A landmark analysis published in Child Abuse & Neglect found that top-down, expert-driven child protection approaches — imposing formal government-managed services — are frequently characterised by low community use and misalignment between formal and informal protection systems. The research argued for community-driven, bottom-up approaches that enable nonformal-formal collaboration, produce greater use of formal services, and achieve higher levels of community ownership and sustainability.
This finding has been replicated across contexts. A systematic review of institutional mechanisms for child protection in India concluded that the child protection framework is strong on paper, but that strengthening child protection committees, decentralising monitoring systems, and providing sufficient resources are important measures that remain largely unrealised.
The UMANG programme in Jharkhand's Godda and Jamtara districts demonstrated what functional Village Level Child Protection Committees can achieve — when civil society organisations actively support them. The programme found that VLCPCs can effectively address issues affecting children when civil society organisations play an active role in activation, when a regular meeting cycle is established, and when they are meaningfully linked to government schemes — producing tangible outcomes on education access, scholarships, and skill development.
What this means for Odisha: The VLCPC structure exists in Odisha under various government frameworks, but is largely dormant in the districts where it matters most. Activating these committees — giving them a specific role, a regular rhythm, and a clear connection to both community concerns and government response mechanisms — is one of the highest-leverage interventions available to child welfare NGOs. This is not glamorous work. It does not generate the impact photographs that fill annual reports. But the evidence suggests it is foundational to everything else.
What BRAC Learned — And Why It Matters for Odisha
BRAC — Bangladesh Rural Advancement Committee — began in 1972 as a small relief organisation and became the world's largest NGO. Much of BRAC's growth was built on a child welfare and education model that is, fifty years later, still one of the most robustly evaluated programmes in global development.
Over 14 million children have graduated from BRAC's pre-primary and primary schools in Bangladesh. The model focuses on enrolling children not served by government schools — particularly children from poor, rural, landless families — in community-based schools with locally recruited teachers.
What made BRAC's model work was not the school buildings — they were minimal. It was the combination of community-based teacher recruitment (women from the same village, trained and supervised consistently), a flexible curriculum calibrated to where children actually were rather than where they "should" be, high female enrolment by design, and an operational model that could survive in contexts where the formal government system could not reach.
BRAC's nutrition and Early Childhood Development work followed a similar logic. An impact evaluation of BRAC's ECD programme used a cluster-randomised controlled trial across 3,120 households and found significant positive effects on child development and nutrition outcomes. The programme model — community-based, built around women in the community as delivery agents, low-cost, and integrated with nutrition support — is directly transferable to the Indian tribal context.
What this means for Odisha: The BRAC lesson is not about replicating Bangladesh. It is about the underlying design principle: that effective child welfare programmes in geographically isolated, high-poverty, low-literacy communities are built around the community itself, not delivered to it from outside. The Anganwadi worker who lives in the village is more valuable than the specialist who visits monthly. The woman who was trained to support breastfeeding in her hamlet is more effective than the hospital nurse who sees mothers once.
Part Three: Five Organisations Doing This Exceptionally Well
1. Gram Vikas (Odisha, India)
Gram Vikas is not famous because it works in an easy context. It is famous because it has made an extraordinary impact in one of the hardest contexts in India — Odisha's most remote and disadvantaged villages — over more than four decades.
Their model — which they call MANTRA (Movement and Action Network for Transformation of Rural Areas) — is built on a non-negotiable principle: universal inclusion. A village does not get a water and sanitation project from Gram Vikas unless every household in the village participates, including Dalit households. This is not idealism. It is programme design. Experience showed that partial coverage produced fractured outcomes and reinforced existing social exclusion.
The child welfare implications of the MANTRA model are profound. Sanitation — specifically the elimination of open defecation — is one of the most powerful predictors of child stunting. India's stubbornly high stunting rates despite improving nutrition are partly explained by the stunting-sanitation link: children who ingest pathogens from contaminated environments cannot absorb nutrients efficiently regardless of what they eat. Gram Vikas's insistence on universal sanitation coverage before programme completion has produced village-level environments where children's nutritional status improves even before direct nutrition interventions are made.
The transferable lesson: Universal inclusion as a design principle, not a aspiration. The most marginalised households — Dalit, PVTG, single-parent — are the ones that standard programmes most often miss. Building the programme around their inclusion from the start, rather than reaching them "when capacity allows," is the structural change that produces equity.
2. BRAC (Bangladesh)
As described above, BRAC's education and early childhood development programmes represent the most robustly evaluated large-scale model for reaching marginalised children with quality education and nutrition support. Two specific practices are directly transferable to Odisha's context.
The first is the community teacher model. BRAC recruits women from the village as teachers — women who are respected locally, who speak the children's language, who can be in the classroom every day because they live there. They receive training and ongoing supervision. This is not a shortcut; it is a deliberate design choice based on the recognition that a formal trained teacher who transfers every two years is less effective than a local woman who is there consistently.
The second is the play-based learning approach for early childhood. BRAC's Play Labs — mobile and community-based early childhood development centres that use play as the primary medium of learning — have been piloted in crisis contexts and low-resource environments with significant results. The model requires minimal infrastructure and can be delivered in an existing community space.
The transferable lesson: For Odisha's Anganwadis to function as genuine early childhood development centres rather than supplementary feeding distribution points, the approach to learning — play-based, child-directed, in the mother tongue — matters as much as the physical infrastructure.
3. Bal Raksha Bharat / Save the Children (India)
Bal Raksha Bharat's community-based child protection work — particularly its approach to activating village-level protection mechanisms and its engagement with child labour communities — has built one of the most operationally grounded models in the Indian context.
The organisation has demonstrated that community-led child protection systems can be effectively established at the village and ward level through NGO facilitation, focusing on prevention and timely response, developing robust community-led child protection systems, and building the agency of children to protect themselves.
Their approach to child labour is particularly instructive for Odisha. Rather than relying on rescue-and-rehabilitation alone — a model that addresses symptoms without addressing causes — Bal Raksha Bharat works at both ends: sensitising trade organisations and businesses to end their use of child labour, while simultaneously supporting families with alternative income sources and connecting children to education.
The transferable lesson: Child labour in Odisha's tribal districts is primarily an economic survival strategy, not a preference. Programmes that address it purely through rescue and legal enforcement, without simultaneously strengthening household economic security and creating genuinely attractive educational alternatives, will see children return to work. The evidence supports an integrated approach.
4. Pradan (India)
Pradan — Professional Assistance for Development Action — has worked in Jharkhand, Odisha, and other tribal states for four decades. Their model of intensive, long-term community immersion — where young professionals spend years embedded in specific communities, building trust and supporting self-help group formation — has produced some of the most documented livelihood and social outcomes in India's tribal belt.
What makes Pradan relevant to child welfare is their understanding of the chain: when tribal women are organised into functional SHGs with access to credit, their household economic security improves; when household economic security improves, children eat more consistently; when children eat more consistently, they grow better and stay in school longer.
Pradan has explicitly documented that the nutritional status of children in SHG households is measurably better than in comparable non-SHG households in the same villages. This is not an accident of selection bias — it reflects the direct link between women's economic agency and child welfare outcomes.
The transferable lesson: Child welfare interventions that do not simultaneously address maternal economic security are working against themselves. In tribal Odisha, where subsistence farming households face months of food insecurity annually, a nutrition programme without a livelihood component is treating a symptom of poverty rather than poverty itself.
5. Pratham (India)
Pratham is most famous for the Annual Status of Education Report (ASER) — the first national assessment of foundational learning outcomes in India, which revealed in 2005 and every year since that millions of Indian children who are enrolled in school cannot read or do basic arithmetic.
Pratham's response to this crisis — the Teaching at the Right Level (TaRL) methodology — has since been validated by multiple randomised controlled trials and adopted by governments across India and Africa. The core insight is deceptively simple: rather than teaching children according to their class level, teach them according to their actual learning level. Group children by what they can do, not by how old they are. Use volunteer community teachers. Dedicate intensive time to foundational literacy and numeracy.
For child welfare in Odisha, Pratham's significance is not just educational. Children who can read by Class 3 are substantially more likely to complete primary school; children who complete primary school are substantially less likely to end up in child labour; children who complete schooling have measurably better health outcomes across their lives. Education is child welfare. The quality of what happens in the classroom — or the Anganwadi — is not a separate sector from nutrition and protection. It is the same child.
The transferable lesson: Odisha's adoption of the NIPUN Bharat Foundational Literacy and Numeracy mission is the right policy direction. What matters now is implementation quality at the block and school level. NGOs can play a crucial role in community monitoring of foundational learning outcomes — not as parallel assessors, but as supporters of teacher quality and community accountability for what children actually know.
Part Four: The Odisha Adaptation — What Transfers, What Doesn't, and What Must Be Built From Scratch
What Transfers Directly
Community-based delivery models work in Odisha's tribal context. The BRAC principle — that local women, trained and supported consistently, are more effective delivery agents than distant specialists who visit periodically — is validated by every successful programme in Odisha's tribal districts. The Anganwadi worker system is a manifestation of this principle, even if it is inconsistently resourced. Programmes that strengthen and support existing community-level workers will consistently outperform programmes that build parallel delivery structures.
Universal inclusion as a design principle is non-negotiable in Odisha's socially stratified tribal villages. Programmes that exclude Dalit households, or that reach Munda communities but not adjacent Bonda villages, replicate and reinforce existing hierarchies. Gram Vikas learned this lesson over decades of work in Odisha itself. It remains the most important design constraint.
Linking nutrition with livelihoods is validated by every successful programme in the Indian tribal context. Pradan, BRAC, and the Bolsa Família model all demonstrate that you cannot sustainably improve children's nutritional status without addressing household food security, which means addressing income and livelihood.
Community ownership of child protection — the bottom-up approach documented in Jharkhand and Sierra Leone — is directly transferable. What it requires is an NGO that is willing to play the role of facilitator and step back, rather than the role of programme implementer who controls the process.
What Requires Significant Adaptation
Cash transfer models require careful adaptation for Odisha's PVTG communities, where financial inclusion is often minimal, banking infrastructure is absent, and literacy levels make navigating documentation requirements difficult. The Mamata scheme has shown what happens when cash transfer models are not adapted: the most marginalised families — PVTG women without Aadhaar, bank accounts, or awareness of their entitlements — are systematically excluded. NGOs working on cash transfers in this context need to build financial literacy alongside the transfer, not assume it.
Formal child protection structures — CWCs, VLCPCs, JJBs — require active NGO facilitation to become functional in tribal Odisha. On paper, the framework exists. In practice, committee members in Malkangiri or Nuapada blocks often don't know their own authority, have no connection to the District Child Protection Unit, and have never handled a referral. Adapting the community-led child protection model for Odisha means investing in this institutional activation — which is slow, unglamorous, and produces outcomes that are difficult to photograph.
Technology-based interventions need to be approached with significant humility. Mobile data collection, digital nutrition monitoring, and tech-enabled early warning systems have tremendous potential in Odisha. But smartphone penetration, connectivity, and digital literacy among frontline workers in PVTG areas remain serious constraints. Programmes designed for Delhi's urban slums will not function unchanged in Malkangiri's forest villages. Any technology deployment needs to be designed with the assumption of intermittent connectivity, low digital literacy, and Odia-language interfaces.
What Must Be Built From Scratch
There are two things that the global evidence consistently points to as important but that essentially do not exist in Odisha's tribal child welfare landscape.
The first is granular, real-time data on child welfare outcomes at the habitation level. Odisha's official nutrition data is collected through the Anganwadi system and aggregated at the block level. This means that a village with a severely malnourished cluster of children can be invisible in official data if the neighbouring villages in the block have better outcomes. Research published in PLOS One using sub-district survey data found that child undernutrition in Odisha exhibits significant spatial heterogeneity at the sub-district level that block-level data completely obscures — and that this heterogeneity cannot be addressed without micro-level evidence. NGOs have the potential to generate and use this data if they are equipped with the right tools — KoboToolbox or ODK for field data collection, mapping software to visualise patterns, and the analytical capacity to turn data into programme decisions.
The second is multi-year, multi-sector programming in the same geographies. Child welfare problems in Odisha's tribal districts are not solved by a three-year nutrition project or a two-year child protection initiative. They require sustained, integrated presence: nutrition, education, child protection, livelihood, water and sanitation, operating together in the same community over long enough periods to produce genuine systems change. Most NGO funding models — annual grants, project-by-project reporting, sector-specific CSR partnerships — work against this. Building the case for five-to-seven-year integrated community investments in Odisha's highest-deprivation blocks is one of the most important advocacy tasks facing the sector.
Part Five: Government Scheme Mapping — What Exists and How NGOs Can Help Communities Use It
The Indian government's child welfare scheme architecture is extensive. The problem is not the existence of schemes — it is the gap between scheme design and last-mile delivery. NGOs that help communities claim what they are already entitled to — without creating parallel delivery systems — provide enormous leverage at relatively low cost.
Integrated Child Development Services (ICDS): The world's largest early childhood programme in terms of coverage, providing supplementary nutrition, immunisation, health check-ups, referral, pre-school education, and nutrition and health education through a network of Anganwadi Centres. NGO role: support Anganwadi quality through community monitoring and building demand for accountability.
POSHAN Abhiyaan (National Nutrition Mission): India's flagship programme to reduce malnutrition, stunting, underweight, and low birth weight across the country. Odisha has its own state implementation plan. NGO role: monitor coverage gaps, support community mobilisation for POSHAN Diwas activities, and document where the scheme is not reaching.
Mamata Scheme (Odisha): ₹5,000 conditional cash transfer to mothers for antenatal and postnatal care compliance. NGO role: enrolment facilitation, documentation support, awareness in PVTG communities, and monitoring of exclusion.
Mission Vatsalya (formerly ICPS): Central government programme for child protection — funds CWCs, JJBs, Child Care Institutions, and District Child Protection Units. NGO role: facilitate activation of VLCPCs, link communities to DCPUs, support CWC referral pathways.
Pada Prusti Karyakram (Odisha): State-specific outreach programme for health and nutrition support in remote tribal areas. NGO role: support awareness and community mobilisation, document coverage gaps in PVTG habitations.
Odisha Millets Mission: While primarily an agricultural programme, its nutrition implications for children are direct — millet-based complementary foods are nutritionally dense and culturally appropriate for tribal communities in Odisha's districts. NGO role: link community nutrition work to millet promotion and procurement.
POCSO Act (2012) and Child Labour (Prohibition and Regulation) Amendment Act (2016): Legal frameworks for child protection and against child labour. NGO role: awareness generation in communities, support for reporting and documentation, linkage with police and legal aid where children are at risk.
Part Six: Further Reading — The Best of What's Been Written
These are not lists of everything. They are the resources that a programme officer or CSR manager should actually read — each annotated with one sentence on why.
Nutrition and the First 1,000 Days:
- Malnutrition and Anemia Among Particularly Vulnerable Tribal Groups of Odisha, India — ICMR-RMRC, Bhubaneswar (2024): The most rigorous recent study of undernutrition among all 13 of Odisha's PVTGs, published in the Indian Journal of Community Medicine. Start here for the Odisha-specific evidence.
- Region Matters: Mapping the Contours of Undernourishment Among Children in Odisha — PLOS One (2022): Demonstrates the sub-district geographic heterogeneity of undernutrition that block-level data misses, and argues for micro-level evidence bases for targeted interventions.
Child Protection Systems:
- Bottom-Up Approaches to Strengthening Child Protection Systems — Wessells, Child Abuse & Neglect (2015): The most influential paper on why top-down child protection systems fail and what community-led approaches achieve differently. Short, rigorous, essential.
- A Guide for Supporting Community-Led Child Protection Processes — Childreninemergencies.org / multiple authors: Practical operational guidance on how to facilitate community-led child protection, based on work across India, Kenya, and Sierra Leone.
Cash Transfers:
- Cash Transfers in the Perinatal Period and Child Welfare System Involvement — Agarwal et al., arXiv (2024): Recent RCT evidence from Flint, Michigan's Rx Kids programme demonstrating that unconditional cash transfers during pregnancy reduce child welfare system involvement among infants.
BRAC and Community-Based Models:
- Primary Education for All: Learning from the BRAC Experience — Quinn et al.: Documents the evolution and evidence base of BRAC's non-formal primary education programme, including cost-effectiveness and gender equity outcomes.
Odisha Context:
- Access to Nutrition in Odisha — Saigal and Shrivastava, Development Policy Review (2022): Examines why nutrition scheme access is restricted for Odisha's most vulnerable groups, focusing on Mamata and the Supplementary Nutrition Programme.
A Final Note: What NGOs Working in This Space Need
The global evidence, taken together, points toward a clear set of conditions that allow child welfare programmes in tribal Odisha to actually work.
Time. The communities where child welfare problems are most acute in Odisha are not accessible — geographically, linguistically, or socially — to programmes that arrive for three years and depart. Trust takes longer than a project cycle to build. Behaviour change takes longer than an output report to document.
Integration. Nutrition without sanitation is fighting with one hand. Child protection without livelihood support is treating symptoms. Education without community health is missing half the child. The evidence consistently shows that integrated, multi-sector programmes in the same geography produce larger effects than the sum of their siloed parts. This is also the hardest thing to fund and to run.
Mother-tongue delivery. Odisha has 62 Scheduled Tribes speaking dozens of distinct languages and dialects. A nutrition education campaign delivered in Odia to a Bonda woman in Malkangiri is not a nutrition education campaign. It is a missed opportunity. Every effective programme reviewed in this Primer — BRAC's schools, Gram Vikas's community meetings, the UMANG protection committees — invested in communication in the community's own language.
Community ownership, not community participation. There is a crucial difference between programmes that involve communities in their implementation and programmes that are genuinely owned and led by communities. The evidence from Sierra Leone, Jharkhand, and Odisha itself is consistent: community-owned programmes are more sustainable, more equitable, and more effective at reaching the most marginalised households. Community participation — where communities execute plans designed by NGOs — looks similar but produces different outcomes.
Data that drives decisions. Too few child welfare programmes in Odisha generate data that actually changes what they do. Community-level data on malnutrition rates, school dropout, and child protection incidents — collected regularly, analysed honestly, and used to correct course — is what separates a programme from a permanent one.
This Sector Primer was written by the JaBaSu Trust knowledge team as part of the JaBaSu Knowledge Commons — a curated library of global and Indian evidence on social sector best practices, contextualised for Odisha.
We update this Primer when significant new evidence emerges. If you find an error, want to suggest an addition, or want to discuss any of the evidence cited here, write to us at knowledge@jabasu.org.
Evidence Grade: B — Multi-study. This Primer draws on multiple independent evaluations and peer-reviewed research from comparable contexts. It does not represent a systematic review.
Last reviewed: April 2026
Related Knowledge Commons Content:
- Practice Note: Community-Based Management of Acute Malnutrition
- Practice Note: Village Level Child Protection Committees — Making the System Work
- Practice Note: Conditional Cash Transfers in India — Evidence and Adaptation
- Org Spotlight: Gram Vikas — Universal Inclusion as Programme Design
- Org Spotlight: BRAC — Community Teachers and the Education Model
- Sector Primer: Health & Nutrition (Sector 08)
- Sector Primer: Education (Sector 04)
Government Schemes Referenced:
- ICDS / Anganwadi System
- POSHAN Abhiyaan
- Mamata Scheme (Odisha)
- Mission Vatsalya
- Pada Prusti Karyakram
- Odisha Millets Mission