The child nutrition study was set in tribal districts in Odisha — Koraput, Rayagada, and Kandhamal — where child stunting and wasting rates remain among the highest in the country despite two decades of targeted nutrition programming. The state runs, as most Indian states do, overlapping systems of social protection: the Public Distribution System, which provides subsidised food grains to Below Poverty Line households, and the Integrated Child Development Services scheme, which provides supplementary nutrition, health check-ups, and preschool education to children under six and pregnant or lactating women.
Standard analytical approaches treat these as separate interventions to be evaluated separately. When we looked at the data across blocks, a different picture emerged.
Households receiving both PDS and ICDS had, in several blocks, worse child nutrition outcomes than comparable households receiving ICDS alone. The result was statistically significant and directionally consistent across the three districts. It was also counterintuitive: two nutrition support programmes, layered together, should produce better outcomes than one.
The mechanism investigation took us to the implementation level.
Beneficiary registration for PDS and ICDS in Odisha operates through separate administrative channels with separate documentation requirements. The ration card system, which governs PDS access, is administered through the Food Supplies and Consumer Welfare Department. ICDS registration, which determines access to supplementary feeding at the anganwadi centre, goes through the Women and Child Development Department. For a household enrolled in both programmes, maintaining active beneficiary status required keeping documentation current in both systems simultaneously.
In the tribal communities we studied, this documentation burden had a specific practical consequence: families who were actively managing compliance with both programmes missed anganwadi supplementary feeding days at higher rates than families enrolled in ICDS alone. The anganwadi requires physical presence for supplementary feeding — it is not a take-home ration in most blocks. On days when families needed to be at the ration shop for PDS collection, or at a government office to resolve a documentation issue with either system, children did not attend the anganwadi. The cumulative caloric loss from missed supplementary feeding days across a six-month period was not negligible.
This is a programme interaction effect of a specific and avoidable kind. The two programmes were not undermining each other in their design intent; they were undermining each other through administrative friction that the design of neither programme had anticipated. The administrative systems were built to operate independently. The households navigating them were not independent of each other.
The finding has a second implication that extends beyond this specific interaction. Convergence — the coordination of multiple social protection programmes to act on the same beneficiary household — is a goal of India's social protection architecture at least in principle. The National Nutrition Mission (POSHAN Abhiyan) explicitly aims to coordinate ICDS with PDS, health, and sanitation programmes. But convergence at the design level does not automatically produce convergence at the implementation level. Two programmes can share a beneficiary list and still generate coordination failures at the point of delivery, because the delivery systems operate through different frontline workers with different supervisory chains and different performance incentives.
Convergence analysis should be built into evaluation design wherever multiple programmes are operating in the same geography. The question is not just whether each programme is working; it is whether the portfolio of programmes is working, and whether the interactions between them are net positive or net negative for the intended beneficiaries.
In Odisha, the specific recommendation was for PDS and ICDS distribution calendars to be coordinated at the block level so that ration distribution and anganwadi supplementary feeding days did not fall on the same day for the same households. This required no policy change, no additional expenditure, and no structural reform. It required an instruction from the District Collector's office. Whether it was implemented is a different question.
District nutrition data from POSHAN Tracker and NFHS-5 Odisha district factsheets. Tribal district analysis supported by Odisha Tribal Empowerment and Livelihoods Programme (OTELP) baseline data. PDS implementation data from DFPD.