English only · Odia translation in progress

Chapter 2: The Paradox of Abundance

In Nabarangpur district, in a village called Tentulikhunti, a concrete storage godown operated by the Primary Agricultural Cooperative Society holds paddy sacks stacked to the ceiling. The 2022-23 kharif marketing season was generous here. The rice came in, was weighed, graded, and receipted. Farmers collected their minimum support price plus the state bonus — over three thousand rupees per quintal, the highest effective paddy price in the district’s memory. The godown is full. The cooperative is functional. The procurement machinery worked.

Three hundred metres from the godown, in the Anganwadi centre that serves the same village, a woman named Sita — this is a composite, but the numbers behind her are not — is thirty-one years old, has three children, and weighs forty-three kilograms. Her haemoglobin count, measured during a Village Health and Nutrition Day screening, was 8.2 grams per decilitre. Anything below 11 is anaemic; below 8 is severely anaemic. She is close to the edge. Her youngest child, a boy of twenty-six months, has been measured three times by the Anganwadi worker. His height-for-age falls below minus two standard deviations on the WHO growth chart. He is stunted. Not visibly emaciated — he eats, he plays, he cries — but his body has quietly economised on growth because the construction materials were not available in the right mix at the right time. The damage is largely irreversible.

The godown is full and the child is stunted. This is not a contradiction. It is the paradox of abundance, and it is the defining structural fact of Odisha’s food economy: a state that produces enough rice to feed itself twice over has 31 per cent of its children stunted and 64 per cent of its women anaemic.

The numbers sit beside each other with the quiet stubbornness of things that should not coexist but do. Rice production: 8 to 10 million tonnes annually. Population: approximately 4.7 crore. Per-capita rice availability from own production: roughly 180 to 200 kilograms per person per year, comfortably above any physiological requirement. Public Distribution System coverage: approximately 3.25 crore beneficiaries, nearly 70 per cent of the state population, receiving rice at zero cost since the integration of PMGKAY into the National Food Security Act framework. Meanwhile: stunting at 31 per cent, wasting at 18.1 per cent, anaemia in children at 64.2 per cent — a figure that rose from 44.6 per cent just five years earlier [NFHS-5 Odisha, 2021; DA&FP Odisha 2022-23; OSCSC Annual Reports; Odisha Economic Survey 2024-25].

The granary is not the problem. Something between the grain and the body is broken. This chapter identifies that something, and names the structural logic that keeps it broken.

The Cross-Domain Lens: The Assurance Game

In game theory, the most famous model is the prisoner’s dilemma: two players, each with an incentive to defect regardless of what the other does, arrive at a mutually bad outcome even though a mutually good outcome exists. The prisoner’s dilemma has dominated policy thinking for decades because it seems to explain so many collective-action failures — pollution, corruption, free-riding. But it is the wrong model for Odisha’s nutrition paradox. In a prisoner’s dilemma, defection is the dominant strategy. No rational individual would cooperate. The only solution is to change the payoffs — through regulation, punishment, or external enforcement.

The assurance game, also called the stag hunt, is structurally different. It was first formalised by Jean-Jacques Rousseau as a parable and later by game theorists including Brian Skyrms. The setup: two hunters can either cooperate to hunt a stag (high payoff, but only if both cooperate) or individually hunt a hare (lower payoff, but guaranteed regardless of what the other does). Unlike the prisoner’s dilemma, cooperation is not irrational — it is, in fact, the payoff-dominant equilibrium. Both players would be better off if both cooperated. The problem is that cooperation requires assurance. If I commit to hunting the stag and you chase a hare instead, I go home empty-handed. The rational response to uncertainty about the other player’s behaviour is to play safe and hunt the hare. This produces the risk-dominant equilibrium: everyone plays safe, everyone gets a suboptimal outcome, and the better equilibrium — which everyone would prefer — is never reached.

The crucial difference: in the prisoner’s dilemma, the cooperative outcome is unstable. In the assurance game, the cooperative outcome is stable — but only if you can get there. The barrier is not incentive but trust. Not selfishness but uncertainty. The players are not defecting because they want to. They are defecting because they cannot be sure the other side will cooperate, and the cost of cooperating alone is catastrophic.

This is the precise structure of Odisha’s nutrition paradox. The cooperative equilibrium exists: every child adequately nourished, every mother healthy, the rice surplus converted into genuine food security. The payoff is unambiguously superior — lower healthcare costs, higher cognitive development, greater labour productivity, reduced welfare spending. Every actor in the system would be better off. But the cooperative equilibrium requires simultaneous action across multiple domains — dietary diversity, clean water, functioning health facilities, effective Anganwadi services, changed intra-household norms, maternal nutrition — and no single actor can afford to invest in their piece without assurance that the other pieces will be in place.

A mother in Nabarangpur could, in theory, spend her limited cash on eggs and green vegetables for her twenty-six-month-old instead of on extra rice. But that investment pays off only if the child also has clean water (so the nutrients are absorbed rather than lost to diarrhoeal disease), a functioning Anganwadi providing micronutrient supplements (so the dietary improvement is not fighting alone against iron and vitamin A deficiency), and a primary health centre within reach (so the inevitable infections are treated before they become growth-faltering episodes). Without assurance on these adjacent conditions, the mother’s rational choice is to maximise the one variable she controls — caloric volume, which means more rice — and accept the suboptimal outcome. She is hunting the hare because she cannot trust the forest to deliver the stag.

The assurance game explains why the paradox persists even when the supply constraint is solved. It is not that anyone is behaving irrationally. Every actor — the mother, the father, the Anganwadi worker, the PDS administrator, the district health officer — is individually rational. The system is collectively irrational because the cooperative equilibrium requires coordinated action and no coordination mechanism exists that provides credible assurance to all players simultaneously.

The Numbers That Don’t Add Up

The paradox begins with an accounting exercise.

On the supply side, the numbers are large and confident. Odisha’s paddy area sits between 4.1 and 4.5 million hectares in a typical kharif season, with another 250,000 to 350,000 hectares in rabi and summer. Total rice production oscillates between 8 and 10 million tonnes. The Odisha State Civil Supplies Corporation procures between 55 and 80 lakh metric tonnes of paddy each kharif marketing season — a volume that places Odisha among India’s four largest procuring states, alongside Punjab, Chhattisgarh, and Telangana. In KMS 2022-23, procurement touched approximately 78 lakh metric tonnes, a reported peak. At a 67 per cent milling ratio, this yields roughly 52 lakh tonnes of rice equivalent from procurement alone [DA&FP Odisha 2022-23; OSCSC Annual Reports; Odisha Economic Survey 2024-25].

The per-capita arithmetic is straightforward. With rice production of 8 to 10 million tonnes and a population of approximately 4.7 crore, per-capita availability from own production sits at 170 to 213 kilograms per person per year. India’s average per-capita rice consumption is approximately 60 kilograms per year (the rest comes from wheat and other cereals). Even accounting for Odisha’s rice-dominant diet, the production surplus is massive.

The PDS architecture matches the surplus. Approximately 3.25 to 3.28 crore beneficiaries are covered under NFSA and state food security schemes — nearly 70 per cent of the total population and close to 80 per cent of the rural population. Priority Household members receive 5 kilograms of foodgrain per person per month; Antyodaya households receive 35 kilograms per household per month. The effective issue price, since the integration of PMGKAY, has been zero. The rice is free. Fair price shops number between 26,000 and 28,000, many operated by women’s self-help groups under Mission Shakti. Electronic point-of-sale authentication covers over 95 per cent of these shops. The system, by every procedural metric, works [Department of Food Supplies & Consumer Welfare, Odisha; NFSA notifications; Ministry of CA,F&PD].

Now the demand side. NFHS-5, conducted in Odisha between June 2019 and January 2020, produced these numbers:

Child stunting (under-5): 31.0 per cent. Down from 34.1 per cent in NFHS-4, down from 45.0 per cent in NFHS-3. Progress — real but achingly slow.
Child wasting (under-5): 18.1 per cent. Essentially unchanged from 19.5 per cent in NFHS-3, fifteen years earlier.
Severe wasting: 6.6 per cent. Up from 5.2 per cent in NFHS-3.
Underweight children: 29.7 per cent.
Anaemia in children 6-59 months: 64.2 per cent. Reversed from 44.6 per cent in NFHS-4 — a catastrophic worsening.
Anaemia in women 15-49: 64.3 per cent. Reversed from 51.0 per cent in NFHS-4.
Children 6-23 months receiving a minimum adequate diet: approximately 10 per cent.
Women with BMI below normal: 20.8 per cent.
Low birth weight: 21.4 per cent of births with recorded weight.

[NFHS-5, India and Odisha State Fact Sheets, IIPS & MoHFW 2021; NFHS-4 Odisha 2017; NFHS-3 Odisha 2007]

Place these two columns side by side. On the left: 8-10 million tonnes of rice, 3.25 crore PDS beneficiaries, free grain, 28,000 fair price shops, 74,000 Anganwadi centres. On the right: one in three children stunted, two in three women anaemic, severe wasting rising, dietary adequacy at 10 per cent, anaemia reversing after a decade of apparent progress.

The numbers do not add up. That is the paradox. And the resolution lies not in finding a missing number but in understanding why caloric surplus and nutritional deficit are not the contradiction they appear to be — they are, in fact, produced by the same structural logic.

The Seven Mechanisms of the Paradox

The surplus-stunting coexistence is not a single failure. It is the product of at least seven mechanisms operating simultaneously, each individually documented, each independently sufficient to produce a significant share of the gap, and each interacting with the others in ways that make the aggregate outcome worse than the sum of parts.

Mechanism 1: Cereal Sufficiency Is Not Nutritional Sufficiency

This is the most fundamental and the most overlooked. Rice is calories. Polished white rice — the form that dominates Odisha’s PDS distribution, market consumption, and household cooking — provides approximately 350 kilocalories and 6-7 grams of low-quality protein per 100 grams. What it does not provide in meaningful quantities is iron (0.6 mg per 100g, against a daily requirement of 17 mg for a lactating woman), zinc (1.2 mg per 100g, against a requirement of 12 mg), vitamin A (zero), vitamin D (zero), vitamin B12 (zero), and folic acid (negligible after milling). A child who eats 300 grams of PDS rice per day gets roughly 1,050 calories and 20 grams of protein. The calories are adequate. The child is still deficient in half the micronutrients required to build bone, blood, brain, and immune function [ICMR-NIN Dietary Guidelines 2020; NNMB Surveys].

The PDS was designed to prevent starvation, not to deliver nutrition. It has succeeded brilliantly at its design objective. Open famine of the kind that killed people in Kalahandi in the 1980s is functionally eliminated. But the design objective and the welfare objective have diverged. The PDS delivers rice. Rice is not nutrition. The gap between the two is the paradox in its most elemental form.

The Comprehensive National Nutrition Survey (CNNS, 2019) measured what rice cannot provide. Iron deficiency accounted for 50-60 per cent of child anaemia cases aged 1-4 years nationally, with higher shares in rural and tribal populations. Odisha’s estimated iron deficiency in children under 5 was 28-34 per cent. But haemoglobin-defined anaemia was 64 per cent — the difference attributable to chronic inflammation, parasitic load, sickle cell trait, and the phytate-binding effect of a cereal-dominated diet that reduces iron bioavailability even from the iron that is consumed [CNNS 2019; Lancet Global Health CNNS analyses].

Mechanism 2: Intra-Household Distribution

A household’s aggregate food acquisition masks what happens inside the household. The rice arrives. The dal is cooked. The question is: who eats first, who eats how much, and who eats the protein.

The pattern is well-documented even if poorly quantified. Men eat first. Children eat second. Women — particularly the woman who cooked — eat last. The consequences are visible in the NFHS data itself: male anaemia in Odisha is 28.5 per cent, female anaemia is 64.3 per cent. The gap is not explained by biology alone. It is explained by a systematic pattern in which women serve and eat the residual — the watered-down dal after the thick portion has been served to the husband, the rice without the fish, the meal without the egg [NFHS-5 Odisha; Harriss-White on intra-household food allocation; IFPRI intra-household distribution research].

The Time Use Survey 2019 provides the temporal evidence for this pattern. Odisha women aged 15-59 spend an average of 301 minutes per day on unpaid domestic services, of which approximately 168 minutes is food-related work. They cook the meal, serve the meal, wash the utensils, and eat whatever remains. As the Women’s Odisha series documented, the woman who produces the food is the residual claimant of the food — an arrangement so normalised that it is invisible even to the women inside it [NSO TUS 2019 Odisha state tables; cross-reference Women’s Odisha full_read series].

The intra-household distribution problem is an assurance game in miniature. A woman who would eat adequately — enough dal, an egg, a second serving — needs assurance that the household’s food budget will stretch to cover everyone. In a household operating at the margin, her eating more means someone else eating less, and that someone is often the earning male whose caloric needs are treated as non-negotiable. She economises on herself. The child, dependent on a depleted mother’s breastmilk and on the same residual-allocation pattern, economises on growth.

Mechanism 3: The Protein Gap

Odisha is a coastal state with a long fishing tradition, extensive aquaculture, and a cultural comfort with non-vegetarian food that distinguishes it from much of north India. Fish, eggs, chicken, and mutton are culturally acceptable in most Odia households outside the Brahmin vegetarian minority. Yet protein consumption remains inadequate across the state’s poor and near-poor populations.

The reasons are economic rather than cultural. Fish prices in rural Odisha have risen faster than general food prices, pushing fresh fish beyond the daily reach of households at the bottom two quintiles. Eggs, at Rs 5-6 each, are affordable in theory but are not purchased daily by most rural households. Milk and dairy consumption is structurally low because Odisha is not a dairy state — unlike Gujarat, Rajasthan, or even neighbouring Jharkhand, per-capita milk availability in Odisha sits well below the national average. Pulses — the vegetarian protein source — have become expensive enough that the dal in a poor household’s dinner is thin, watered, and shared across more people than its protein content can support [NFHS-5 dietary diversity indicators; HCES 2022-23 summary; NSSO 68th Round consumption data].

The dietary diversity indicators from NFHS-5 tell the story. Only about 10 per cent of children aged 6-23 months in Odisha received a minimum adequate diet — a composite of minimum meal frequency, minimum dietary diversity, and minimum milk feeds. This is not a supply failure. It is a composition failure. The calories are arriving in the form of rice. The protein, iron, zinc, and vitamins that would convert those calories into growth are not arriving in sufficient quantity or variety.

Mechanism 4: The WASH Channel

A child can eat an adequate meal and still fail to grow if the gut cannot absorb the nutrients. This is the environmental enteropathy pathway, and it may account for 25 to 40 per cent of stunting in Indian contexts.

The mechanism is physiologically precise. Chronic exposure to faecal bacteria through contaminated water, soil, and food creates persistent inflammation in the small intestine — environmental enteric dysfunction. The inflamed gut diverts metabolic energy from growth to immune defence and reduces the absorption efficiency for iron, zinc, and other micronutrients. A child subjected to repeated episodes of diarrhoeal disease experiences cumulative growth faltering that compounds month by month through the critical first thousand days [Humphrey, Lancet 2009; Spears and Coffey, Where India Goes; Lancet WASH Series 2014].

Odisha’s sanitation trajectory has improved dramatically under Swachh Bharat Mission. Before SBM, household latrine coverage in rural Odisha was below 20 per cent. By 2019, the state was declared Open Defecation Free. But the gap between construction and usage is documented: NFHS-5 reports approximately 75 per cent of households using improved sanitation facilities, against the 100 per cent claimed at the construction level. In tribal districts — precisely where stunting is highest — the usage gap is wider. A latrine that exists but is not used by all household members does not break the faecal-oral transmission pathway [NFHS-5 Odisha; IDInsight SBM studies; cross-reference Environmental Odisha full_read series].

Water quality compounds the problem. Jal Jeevan Mission has expanded household tap connections, but coverage in tribal interior villages remains incomplete. Diarrhoeal disease affected 7 per cent of under-5 children in the two weeks preceding the NFHS-5 survey, and only 60-65 per cent of those received ORS and zinc treatment. Each untreated episode depletes nutrient stores and contributes to the cumulative insult that produces stunting.

Mechanism 5: The ICDS Delivery Gap

The Integrated Child Development Services infrastructure in Odisha is large: approximately 74,000 Anganwadi centres, 71,000 Anganwadi workers, and coverage of 35-40 lakh children aged 6 months to 6 years. The architecture is correctly designed — a distributed service network with defined deliverables (supplementary nutrition, growth monitoring, immunisation referrals, nutrition counselling, pre-school education). The implementation is unevenly broken.

The gaps cluster where they matter most. Approximately 25-30 per cent of Anganwadi centres operate from rented or inadequate buildings. Take Home Ration supply chains experience one-to-three-month stock-outs intermittently in tribal blocks during monsoon. Growth monitoring data is recorded inconsistently, undermining the system’s capacity to identify and refer severely malnourished children. Worker honorarium delays have triggered protests in 2021 and 2023. The POSHAN Tracker captures delivery — THR distributed, meals served, home visits conducted — but not outcomes. An Anganwadi worker records that thirty packets of chhatua mix were distributed. Whether the child ate the mix, whether the household shared it among four children instead of giving it to the target beneficiary, whether the child’s gut absorbed the iron in the mix — none of this is visible to the system [CAG ICDS Audit Odisha; WCD Odisha Annual Reports; POSHAN Tracker dashboards; EPW on POSHAN Tracker gaps].

The implementation variance is itself a paradox. Khordha and Cuttack Anganwadis report 90 per cent-plus beneficiary coverage and high attendance. Nabarangpur and Malkangiri Anganwadis report lower attendance, longer distances from remote hamlets, and higher sevika vacancies. The centres that most need to be high-functioning are structurally the hardest to staff, supply, and monitor. This is a systems pattern documented across the Institutional Design series: infrastructure built for the average fails at the margins where it is needed most.

Mechanism 6: The Maternal Loop

Anaemic mothers produce low-birth-weight babies. Low-birth-weight babies enter the stunting pathway from month zero. The intergenerational transmission of undernutrition is the most structurally entrenched mechanism of the paradox — and the hardest to interrupt because it operates across the lifespan, not within a single programme cycle.

The numbers trace the loop. NFHS-5 reports that 64.3 per cent of Odisha women aged 15-49 are anaemic — up from 51 per cent five years earlier. Among pregnant women, 61.8 per cent are anaemic. Twenty per cent of women have BMI below 18.5. The Mamata scheme provides conditional cash transfers; institutional delivery has reached 92 per cent; antenatal care coverage has improved to 70.5 per cent. Yet low birth weight remains at 21.4 per cent and the anaemia figures have worsened.

The maternal loop operates through a biological mechanism that policy cannot shortcut. A girl who was stunted at age two becomes an underweight adolescent. Adolescent girl anaemia in Odisha is 65.5 per cent. She marries — median age 20.4 years, with 19.8 per cent still married before 18. She becomes pregnant while anaemic and under-BMI. The pregnancy depletes her further. She delivers a baby whose fetal programming has already been compromised. The baby enters the Anganwadi system at six months — by which time six months of the critical thousand-day window have already passed in a suboptimal nutritional environment. The Anganwadi provides chhatua. The PDS provides rice. The loop does not break [NFHS-5 Odisha; Barker 1998; Lancet Maternal and Child Undernutrition Series 2013; UNICEF 1000 Days framework].

This is the assurance game at its most vicious. A girl’s body cannot invest in growth without assurance that the inputs (iron, protein, clean water, healthcare) will be there across the entire developmental window from conception through adolescence. The inputs arrive piecemeal, late, and incomplete. The body, unable to trust the supply chain, makes the metabolic choice to survive rather than thrive.

Mechanism 7: The Knowledge-Action Gap

Even where nutritional knowledge exists, social norms and time poverty prevent women from acting on it.

The Anganwadi worker conducts nutrition counselling sessions. The messages are clear: exclusive breastfeeding for six months, introduce complementary foods at six months, include eggs and green vegetables, feed children before adults. The mother knows these things. She may even believe them. But she is also the woman who wakes at five, collects water, cooks breakfast, serves her husband, feeds the older children, works in the field or walks to an MGNREGA site, returns, collects firewood, cooks dinner, serves everyone, eats the residual, and sleeps. Her day contains no slack for the additional food preparation, market visits, and feeding attention that adequate child nutrition requires.

The Time Use Survey evidence from the Women’s Kitchen Labor research is precise here: rural Odisha women spend an average of 180-190 minutes per day on food preparation alone, with additional hours on fuel collection and water fetching. A mother who knows she should feed her toddler mashed egg and spinach separately from the family rice-dal meal faces a time-cost and a fuel-cost for the additional cooking. In a household with a single chulha and no refrigeration, preparing a separate child meal is not a matter of knowledge but of logistics.

Social norms compound time poverty. In many joint-family households, the mother-in-law controls the kitchen, the meal composition, and the feeding order. A young mother who wants to feed her child differently must negotiate with an authority structure that may not share her nutritional priorities. The knowledge-action gap is not an information problem. It is a power problem operating through the medium of food.

The Tribal Belt: Where the Paradox is Sharpest

The KBK+ districts — Koraput, Kalahandi, Bolangir, Nabarangpur, Malkangiri, Rayagada, Nuapada, Sonepur, plus Kandhamal and Gajapati — are the persistent poverty core of Odisha and the place where the paradox cuts deepest.

NFHS-5 district-level data reveals the scale. Nabarangpur: stunting at 42-46 per cent. Malkangiri: 40-44 per cent. Koraput: 38-42 per cent. Rayagada: 36-40 per cent. Kandhamal: 35-39 per cent. Kalahandi: 34-38 per cent. Child anaemia in these districts exceeds 65-72 per cent. The Nabarangpur-Khordha gap on stunting is approximately 20 percentage points — same state, same PDS, same ICDS, same government [NFHS-5 District Fact Sheets, Odisha 2021].

By social group, Scheduled Tribe children show stunting at 37-40 per cent, compared to 23-27 per cent for the general category. The ST-General gap on stunting — roughly 13-14 percentage points — is as large as the gap between Odisha and Kerala. Odisha’s tribal population is 22.8 per cent of the state (Census 2011), concentrated precisely in the districts where the numbers are worst.

The tribal belt nutrition collapse has a specific structural cause that the seven mechanisms above do not fully capture: the loss of forest food.

Tribal communities in Koraput, Rayagada, Kandhamal, and Mayurbhanj historically drew significant nutrition from a diverse forest food base — wild tubers (tikkor, jungli alu), leafy greens (koilari, koinar, saag varieties), fruits (mango, tamarind, mahua, char, kendu), mushrooms, honey, stream and pond fish, small game, insects, and ants. ICAR-CRIDA studies in Koraput found that wild greens contributed 15-25 per cent of monsoon-season micronutrient intake in households that still foraged; over 100 edible wild plant species are documented in the tribal belt [ICAR-CRIDA Koraput; Debal Deb; Agragamee and Living Farms Odisha].

This food base has collapsed through several reinforcing pressures: shifting-cultivation bans under forest conservation rules; Forest Rights Act implementation failure (the Tribal Odisha series documented a 72 per cent rejection rate on community forest rights claims); land alienation from mining, dams, and non-tribal acquisition; the elimination of millets from the cropping system as paddy-focused extension and procurement incentives took hold; and the quiet, progressive substitution of a nutritionally diverse traditional diet with PDS rice.

The substitution effect is the hidden cost of the procurement machine’s success. A tribal household that once ate ragi porridge with wild greens and occasional fish now eats PDS rice with thin dal. Open hunger — the kind that made Kalahandi a national shame in the 1980s — has been eliminated. But dietary diversity has narrowed catastrophically. The household has moved from potential caloric scarcity plus nutritional adequacy to guaranteed caloric sufficiency plus nutritional deficit. The rice arrived. The iron, zinc, vitamin A, and protein left. The aggregate metric (food security) improved. The disaggregated metric (nutritional security) worsened.

The sickle cell dimension compounds this in specific tribal populations. Sickle cell allele frequency in some Odisha ST populations — notably in Sundargarh, Mayurbhanj, and Keonjhar — reaches 10-15 per cent carrier and 1-3 per cent homozygous disease. Sickle cell disease causes chronic haemolytic anaemia, compounds nutritional anaemia, complicates iron supplementation (because the anaemia is not primarily iron-responsive), and reduces growth potential independent of nutrition. A tribal child’s anaemia is not one problem. It is iron deficiency plus phytate inhibition from the cereal-dominated diet plus parasitic infection from poor water plus, in some cases, sickle cell trait. Any single-factor intervention is fighting a multi-factor problem [ICMR Sickle Cell Mission; NSCAEM 2023; ICMR Tribal Health Expert Committee Report 2018].

The Odisha Millets Mission, launched in 2017-18, represents the most serious policy counter-effort. It has reintroduced ragi (finger millet) into tribal cropping systems, ICDS supplementation (ragi ladoos and ragi porridge in Anganwadis across tribal districts), and PDS procurement. By 2023-24 the mission had expanded to 15-19 districts and over 1,400 panchayats, with ragi procurement through the existing paddy procurement machinery. WASSAN evaluations in pilot districts have reported modest but positive anthropometric improvement where ragi was consistently delivered. But the mission remains small relative to the paddy economy — ragi procurement is measured in tens of thousands of tonnes against paddy procurement in tens of millions. It is a corrective, not yet a transformation [Odisha Millets Mission Annual Reports; WASSAN OMM Evaluation; ICRISAT millet supplementation studies; Down to Earth].

The Dongria Kondh of the Niyamgiri hills offer a counter-example that sharpens the analysis. Nutritional studies during and after the Niyamgiri resistance documented that the Dongria diet, while calorically modest, maintains dietary diversity through seasonal foraging. Dongria child anthropometrics are comparable to or slightly better than surrounding tribal populations on stunting. The hypothesis — contested but not rejected — is that preservation of the traditional food system protects against the dietary diversity collapse seen where PDS substitution has gone furthest. The forest food was not a primitive backup. It was the nutritional system. The rice is the backup — and it backs up only calories, not nutrition.

The Assurance Problem in Practice

Return now to the game theory frame and watch how each mechanism is, at its core, a failure of assurance.

The mother’s calculation. A mother in Nabarangpur who would feed her child two eggs per week instead of extra rice can afford the Rs 10-12 per week. But the eggs deliver their nutritional payoff only if the child also has clean water (otherwise diarrhoeal disease washes the protein out), a functioning health centre within reach (otherwise the next infection becomes a growth-faltering episode), and an Anganwadi providing consistent micronutrient supplementation (otherwise the eggs fight alone against a multi-front deficiency). Without assurance on these complementary inputs, the egg investment is high-risk. The hare — more rice, guaranteed calories, no uncertainty — is the safer bet. She hunts the hare.

The father’s calculation. A man who would share the fish or the egg equally with his wife and children — rather than eating first and most, as norm dictates — needs assurance that his own caloric needs will be met without the protein priority. In a household where the earning male performs heavy physical labour (paddy transplanting, construction, MGNREGA earthwork), his caloric requirement is genuinely high. He is not being selfish in the simple sense; he is allocating the scarce high-quality food to the body whose labour income the household depends on. The assurance he lacks is that the household food budget will stretch to adequate protein for everyone. It will not, because the budget is built around free rice and purchased everything-else, and the everything-else is where the money runs out.

The Anganwadi worker’s calculation. A sevika who would spend her time on intensive growth monitoring and nutrition counselling — tracking each child’s weight gain, conducting home visits, counselling mothers on complementary feeding — can do so only if her supply chain delivers the Take Home Ration on schedule, her supervisor supports her rather than burdening her with additional administrative tasks, and her honorarium arrives without the two-month delays that have become routine. Without these assurances, she optimises for what the system measures: distribution counts, registers filled, meetings held. The POSHAN Tracker captures her compliance. It does not capture her impact. She hunts the hare.

The PDS administrator’s calculation. A district food supplies officer who would diversify the PDS beyond rice — adding pulses, cooking oil, fortified flour, eggs — needs assurance of cold chain infrastructure (for eggs), reliable procurement at scale (for pulses at consistent quality), and political cover for a change that disrupts the established rice-procurement-milling chain where powerful interests operate. The rice system works. It is not nutritionally adequate, but it is logistically proven, politically safe, and fiscally manageable because the central government bears most of the cost. Diversification is the stag. Rice is the hare. The administrator hunts the hare.

The policy designer’s calculation. A state nutrition officer who would design an integrated convergence programme — linking ICDS, PDS, WASH, health, and women’s empowerment into a single household-level intervention — needs assurance that the five departments involved will actually coordinate, that the separate budget lines will be fungible, that the separate MIS systems will share data, and that the political leadership will sustain the programme across election cycles. Every department runs its own schemes, its own field staff, its own budget cycle, and its own reporting calendar. The convergence is the stag. The departmental silo is the hare. The officer designs for the silo.

At every level of the system, from the household to the state secretariat, the payoff-dominant equilibrium — coordinated, integrated, nutrition-focused action — exists. Everyone knows it exists. The POSHAN Abhiyaan was explicitly designed to move toward it, introducing district nutrition committees, convergence action plans, and monthly cross-departmental meetings. In practice, the convergence mechanisms function with variable intensity. In aspirational districts, some concrete joint action occurs. In most districts, the monthly meetings happen but the data sharing, joint targeting, and coordinated delivery remain weak. The stag hunt fails because no player has enough trust that the others will cooperate to justify the risk of cooperating alone.

What Closed the Gap Elsewhere

The paradox is not universal. Other states and countries have broken it or substantially reduced it. Their paths illuminate what Odisha would need to do.

Kerala: The Substrate Model

Kerala’s stunting rate stands at 23.4 per cent — nearly 8 percentage points below Odisha’s. The difference is not primarily a nutrition programme story. It is a cumulative outcome of fifty years of investment in the substrate on which nutrition programmes operate.

Female literacy in Kerala exceeds 92 per cent — the highest in India, against Odisha’s 64 per cent. The downstream effect is that educated mothers make better feeding decisions, negotiate more effectively within the household, and access health services earlier. Kerala’s public primary healthcare system, built over generations, catches growth faltering and refers children before the damage becomes irreversible. Decentralised local self-government — panchayati raj with substantial financial and administrative powers — creates accountability for last-mile service delivery that Odisha’s more centralised system cannot match. Low fertility and late marriage (median age at first marriage above 22 for women) produce more spaced pregnancies and healthier mothers. The dietary base is more diverse — fish, coconut, banana, tapioca, vegetables — supported by commercial food markets that bring variety to even small-town consumers [Dreze and Sen, India: Development and Participation; NFHS-5 Kerala; Kerala Development Model literature].

Kerala’s ICDS and midday meal systems run on top of these foundations. They do not carry the load alone. The assurance problem is solved not by any single programme but by the substrate: a mother in Kerala who feeds her child adequately has reasonable assurance that the school will teach, the clinic will treat, the water will be clean, and the toilet will be used. The cooperative equilibrium has been reached because the adjacent systems provide credible assurance.

Tamil Nadu: The Institutional Depth Model

Tamil Nadu’s stunting rate is 25 per cent. The path to that number is different from Kerala’s but equally instructive.

Tamil Nadu built the world’s most mature integrated nutrition system through the Tamil Nadu Integrated Nutrition Project (TINP), launched in the 1980s with World Bank support. The Noon Meal Scheme, started by M.G. Ramachandran in 1982, predates the national Mid-Day Meal programme by over a decade. The scheme includes an egg in every school meal on most days — a protein intervention that Odisha has adopted partially but not at Tamil Nadu’s consistency and scale. Tamil Nadu’s ICDS operates at higher intensity per child: better Anganwadi buildings, higher supervisor density, more consistent supply chains.

But the deeper difference is institutional continuity. The noon meal scheme has survived every government change in Tamil Nadu since 1982 because both DMK and AIADMK treated it as electorally sacrosanct. Forty-plus years of iteration have produced an institutional depth — trained staff, established supply chains, tested menus, community expectations — that cannot be replicated by a five-year programme cycle. Odisha’s PM POSHAN implementation is roughly fifteen years younger in institutional age. The gap is not design; it is depth [World Bank TINP evaluations; EPW on Tamil Nadu nutrition; Dreze-Sen].

Bangladesh: The Women’s Agency Model

Bangladesh has reduced stunting faster than any South Asian country over the past three decades — from over 60 per cent in the early 1990s to approximately 28 per cent by 2019. The reduction happened at lower per-capita income than Odisha’s, demolishing the argument that income is the binding constraint.

The drivers were: rapid improvement in female education and workforce participation (garment manufacturing brought millions of women into the formal economy); widespread adoption of oral rehydration therapy and immunisation (BRAC and Grameen created a parallel delivery infrastructure); microfinance-linked women’s empowerment that gave women control over household spending, including food spending; and targeted supplementary feeding programmes that reached children in the critical window [BDHS/NIPORT reports; Lancet Bangladesh nutrition series; World Bank Bangladesh nutrition reviews; BRAC evaluations].

The Bangladesh lesson for the assurance game is precise: what broke the risk-dominant equilibrium was not a single programme but a shift in the agency of the player who controls intra-household food allocation — the mother. When women earned income, controlled budgets, had education to translate knowledge into action, and had access to health services through a parallel NGO infrastructure, the assurance problem was solved at the household level. The mother did not need to trust the government’s convergence plan. She had enough control to make the investment in her child’s nutrition herself.

What Would It Take for Odisha?

The comparison reveals three pathways, all demanding, none quick.

First, the substrate path (Kerala): invest massively in female education, primary healthcare, sanitation, and local governance over decades. This is the highest-return, longest-horizon strategy. It does not fit in a five-year POSHAN Abhiyaan cycle.

Second, the institutional depth path (Tamil Nadu): pick one or two nutrition interventions — school meals with daily eggs, integrated Anganwadi services with consistent supply chains — and sustain them with bipartisan commitment across multiple government cycles until they reach institutional maturity. This requires political will that transcends party, which Odisha has demonstrated with OSDMA (a single institution that works, sustained across government changes) but has not replicated in nutrition.

Third, the women’s agency path (Bangladesh): use women’s economic empowerment — through Mission Shakti, SHG employment, skill development — as the vehicle for shifting intra-household food allocation. When the mother earns, controls, and decides, the assurance problem is partially solved at the source. Mission Shakti has the architecture; the question is whether it reaches deeply enough into the poorest, most remote households where the paradox is sharpest.

None of these paths is a programme. All of them are systems changes. And that is the nature of the assurance game: you cannot solve it with a single intervention. You solve it by changing the conditions under which every player operates, until the cooperative equilibrium becomes not just desirable but safe to choose.

The Nutritional Transition: The Second Paradox

While rural and tribal Odisha remain locked inside the deficit phase of nutrition, urban Odisha is developing the metabolic signature of the surplus phase. The same state, the same decade, contains both stunting and diabetes as mass phenomena.

NFHS-5 reports that 20.3 per cent of Odisha women aged 15-49 are overweight or obese (BMI at or above 25.0), up from 16.5 per cent in NFHS-4. In urban Odisha, the overweight rate among women reaches approximately 30-35 per cent. Men’s overweight sits at 17.3 per cent statewide and higher in cities. Bhubaneswar and Cuttack are experiencing rapid expansion of packaged food, sugar-sweetened beverages, refined carbohydrate convenience foods, and ultra-processed snacks. The traditional Odia diet — rice, dal, vegetables, fish, moderate oil — is being displaced in middle-class urban households by a diet whose caloric density has risen while its micronutrient density has fallen [NFHS-5 Odisha; HCES 2022-23 summary; Popkin on nutrition transition].

The ICMR-INDIAB study places Odisha’s diabetes prevalence at approximately 8-10 per cent in urban adults and 3-5 per cent in rural adults, with prediabetes significantly higher in both groups. These numbers are lower than Tamil Nadu or Kerala’s urban diabetes rates but rising, and they carry a particular danger rooted in the paradox itself.

The danger is the Barker hypothesis made population-wide. A mother who was undernourished in her own childhood — stunted, iron-deficient, metabolically programmed for scarcity — produces a baby whose fetal programming emphasises fat storage and insulin resistance. This is an adaptive response: a body expecting scarcity hoards every calorie. But when that child, programmed for scarcity, encounters the urban obesogenic environment of Bhubaneswar in the 2020s — the Maggi noodles, the Frooti, the packaged biscuits, the refined rice and oil — the metabolic vulnerability manifests as early-onset diabetes, central obesity, and cardiovascular risk at lower BMI than would be expected in populations without the undernutrition history [Barker 1998; Yajnik on the thin-fat Indian phenotype; Fall CHD Indian cohort studies; New Delhi Birth Cohort].

This is the second paradox of abundance: the transition from caloric deficit to metabolic disease without ever passing through nutritional adequacy. Rural Odisha is still fighting stunting. Urban Odisha is developing diabetes. And the two are biologically continuous — the stunted rural child who migrates to the city is metabolically primed for the diseases of surplus.

The data on the double burden is emerging but already alarming. Hypertension affects approximately 22-25 per cent of Odisha’s adult population, higher in urban areas. Central obesity in urban women exceeds 35-40 per cent by NFHS-5 measures. Women with high random blood glucose have risen from approximately 5-6 per cent to 7-8 per cent between survey rounds. And the metabolic syndrome — the cluster of high blood pressure, high blood sugar, excess waist circumference, and abnormal lipid levels that predicts cardiovascular disease — is estimated at 25-30 per cent in urban Odisha women by regional studies [NFHS-5 Odisha; ICMR-INDIAB; Indian Heart Journal on NCDs in eastern India].

The state’s public health system, already struggling to close the child nutrition gap, will simultaneously face rising diabetes, hypertension, and cardiovascular disease. The National Programme for Prevention and Control of Non-Communicable Diseases operates NCD clinics at district and community health centre levels, but the treatment cascade — from identification to treatment initiation to adherence to control — remains weak, with drop-off at each stage. Two agendas — undernutrition and NCDs — pulled from different policy frameworks, funded by different budgets, delivered by different staff, though they are biologically the same story told at different points in the lifecycle.

Connections

The paradox of abundance does not sit within the food domain alone. It connects to patterns identified across prior series in ways that are worth naming explicitly.

The extraction equilibrium identified in The Long Arc series — the structural pattern in which minerals and talent leave Odisha while the value is captured elsewhere, with welfare substituting for development — has a precise nutritional analogue. Odisha’s rice economy extracts nutritional wealth from households the same way the mineral economy extracts geological wealth from the state. The farmer grows the rice, sells it at MSP, receives cash. The rice enters the procurement system, is milled, dispatched to the central pool or the PDS. The household eats rice. The cash buys more rice. The protein, micronutrients, and dietary diversity that would convert the rice into health are not procured because the procurement system procures rice and nothing else. The nutritional value — like the mineral value — is added elsewhere or not added at all. The household, like the state, sits at the bottom of the value chain of its own production.

The intra-household distribution mechanism connects directly to the Women’s Odisha series and its documentation of gendered labour, gendered eating, and gendered health outcomes. The shadow GDP of women’s kitchen labour in Odisha — estimated at Rs 3.2 to 6.8 lakh crore per year at minimum-wage valuation, roughly half the state’s measured GDP — is produced by the same women who eat last. The largest food production system in the state is operated by a workforce that is systematically malnourished. This is not irony; it is the structural logic of a system in which women’s labour is treated as infinitely available and women’s nutrition is treated as infinitely compressible.

Delhi’s Odisha series documented the PDS as a central policy instrument — a national food security architecture designed in Delhi, funded by Delhi, and imposed on states as a one-size-fits-all caloric delivery mechanism. The PDS provides rice because rice is what the central procurement system procures. It does not provide eggs, pulses, cooking oil, or fortified foods at meaningful scale because the central architecture was not designed for dietary diversity. States that have diversified their PDS — Tamil Nadu’s universal system with wheat and pulses, Chhattisgarh’s own food security act — have done so through state initiative and state funding. Odisha’s PDS is generous on coverage and on price, but it inherits the central architecture’s structural limitation: it solves for calories, not for nutrition.

The tribal belt nutrition collapse connects to the Tribal Odisha series and its documentation of constitutional betrayal: PESA violations, Forest Rights Act implementation failure, land alienation, the systematic dismantling of the autonomous food systems that sustained tribal communities for centuries. The forest food loss is not an accident. It is the nutritional consequence of a political process — the same process that produced 136 CAG-documented violations of PESA in Odisha — in which tribal rights to land, forest, and food were legally guaranteed and practically denied.

Honest Limitations

The assurance game is a clarifying frame, but it carries two risks that should be stated plainly.

First, the frame risks making the problem sound like a coordination failure that could be solved by “just providing assurance” — perhaps through a better convergence plan, a more integrated MIS, a district-level nutrition committee that actually meets. The reality is messier. The structural barriers that prevent the cooperative equilibrium from being reached — caste hierarchies that determine who eats first, gender norms that determine who eats last, poverty that compresses every household budget to the bone, state capacity gaps that leave Anganwadis understaffed and health centres unreachable — are not merely coordination problems. They are power problems. The assurance game assumes players who want to cooperate but lack trust. In many households and institutions, the issue is not that cooperation is desired but risky. The issue is that some players benefit from the current arrangement. The man who eats first is not failing to coordinate. He is exercising a power that the household structure grants him. The milling lobby that resists PDS diversification beyond rice is not lacking assurance. It is protecting a rent. The assurance game illuminates the structure of the paradox but does not fully explain the politics that sustain it.

Second, the frame risks implying that the paradox is solvable within Odisha’s current fiscal and institutional constraints if only the coordination were improved. The comparator evidence suggests otherwise. Kerala took fifty years of sustained investment in female education and public health to reach its current nutritional outcomes. Tamil Nadu has had four decades of institutional depth in school meals. Bangladesh’s transformation required a parallel NGO infrastructure (BRAC, Grameen) that has no equivalent in Odisha. These are not five-year programme cycles. They are generational commitments. Framing the problem as an assurance game clarifies what needs to change. It does not make the change fast, cheap, or easy.

Closing Synthesis

Return to Tentulikhunti. The godown is still full. The child is still stunted.

The distance between the two — three hundred metres of village road — is the distance across which the paradox operates. On one side, a system that works: procurement centres, MSP, state bonus, OSCSC, milling contracts, PDS dispatch, FPS authentication, rice delivered. The system was designed to move grain from field to household, and it moves grain from field to household. It is a triumph of logistics.

On the other side, a body that does not grow. Not because the grain is absent but because the grain is not nutrition. Because the household that receives the grain allocates it on patterns of gender and age that leave the most vulnerable members — the pregnant mother, the toddler — as residual claimants. Because the water the child drinks carries pathogens that inflame the gut and steal the nutrients the grain provides. Because the Anganwadi that should supplement the household diet is intermittently stocked, inconsistently staffed, and measured on distribution rather than outcomes. Because the mother who cooks the meal, serves the meal, and eats what remains is herself anaemic, and her anaemia will be transmitted to the next child through the relentless loop of biology.

The seven mechanisms are individually documented. The assurance game shows why they persist together. Each actor is rational. The system is irrational. The cooperative equilibrium — in which every input arrives simultaneously, every household allocates food equitably, every child receives the full complement of nutrients in a body with a functioning gut in a village with clean water served by a staffed Anganwadi near a functioning health centre — that equilibrium exists. It is payoff-dominant. Everyone would be better off. But no single player can afford to move toward it without assurance that all the others will move too.

What would it take for both the granary and the children to be full? Not more rice. Not even, by itself, more money. It would take the thing that game theory says is hardest to produce and most valuable when present: credible assurance. Assurance that the ICDS will deliver consistently. Assurance that the water will be clean. Assurance that the health centre will function. Assurance that the mother’s investment in her child’s dietary diversity will not be wasted by the next episode of diarrhoea in an unmonitored village. Assurance that the egg she buys will be absorbed by a healthy gut, treated by an available nurse, supplemented by a stocked Anganwadi, sustained by a mother who herself was adequately nourished.

OSDMA proved that Odisha can build institutions that provide credible assurance — in disaster management, where the assurance is that the warning will come, the shelter will be open, the relief will arrive. The question is whether the same institutional logic can be applied to the slow disaster of malnutrition, where the warning is a growth chart, the shelter is a functioning Anganwadi, and the relief is a diet that contains not just calories but the full set of materials a human body needs to grow.

The granary is full. The paradox is that fullness, by itself, feeds no one. What feeds people is not the grain in the godown but the system that converts grain into nourishment — and that system, in Odisha, is broken at every joint between production and the body. Fixing it is not a procurement problem, not a logistics problem, not even primarily a funding problem. It is a trust problem. An assurance problem. A coordination problem operating through the deepest structures of gender, caste, poverty, and state capacity.

The stag is there. The forest is there. No one hunts it because no one is sure the others will show up.