Abstract

Food insecurity remains a persistent challenge in rural India, shaped by interactions between agricultural production, livelihoods, caste, gender, and access to public distribution systems. This paper examines its determinants using the Food Insecurity Experience Scale (FIES) developed by FAO. The analysis draws on primary surveys conducted in purposively selected villages across six Indian states between 2015 and 2023. These surveys captured household socio-economic conditions, focusing on agriculture, land, food consumption, common property resources, and social protection.

Household questionnaires collected data on demographics, education, occupation, land and cropping, food consumption, indebtedness, housing, sanitation, and welfare access. The FIES module, translated into local languages, was administered to one randomly selected adult per household.

An ordered logit regression model identified determinants of food insecurity using FIES-based probability scores. The dependent variable comprised nine ordered categories of severity. Independent variables included per capita cereal production, occupation, social group (with emphasis on Scheduled Tribes and Scheduled Castes), gender of respondent, and access to the Public Distribution System (PDS).

Results show that higher per capita cereal production significantly reduces the probability of moderate and severe food insecurity, underscoring the importance of land and agriculture. Stable, assured incomes play a protective role. Households dependent on casual manual labor, pensions, or remittances were most vulnerable, while those with land, businesses, or private salaried jobs did not differ significantly from government employees.

Caste disparities persist: Scheduled Tribe households were significantly more food insecure than non-SC/ST households, even after controlling for occupation and land, pointing to structural disadvantage. Gender differences were not statistically significant, though patterns are suggestive. In poorer occupational groups, men appeared more food insecure, possibly due to migration for precarious informal work. In wealthier or salaried households, women appeared more vulnerable, potentially reflecting intra-household inequalities in food allocation and reporting.

Access to subsidized food through the PDS showed a positive association with food insecurity, reflecting its targeted design, with ration cards concentrated among vulnerable households. Two caveats apply: many food-insecure households still lack ration cards, and subsidized grain is often insufficient to ensure food security. This highlights structural limitations in current PDS design and coverage.

Overall, food insecurity in rural India is shaped by agricultural access, occupation, and social identity, with land and secure incomes central. Scheduled Tribe households and those reliant on manual labor or remittances remain especially vulnerable. While the PDS reaches many insecure households, it is insufficient to eliminate food insecurity, indicating the need for broader entitlements. This study contributes rare primary evidence across diverse agro-ecological and social contexts and links household data with internationally comparable measures, reinforcing the importance of agriculture, regular employment, and addressing structural inequalities across caste and occupation.