The entitlement approach is Sen's analytical framework for understanding famines and other catastrophes of deprivation. Developed through his empirical study of the Bengal famine of 1943, it demonstrates that such catastrophes typically result not from absolute scarcity of the relevant resource but from the collapse of the institutional and economic mechanisms — entitlements — that determine who can access the resource that exists. Bengal had enough food in 1943. People starved because wartime inflation, speculative hoarding, distribution failures, and the suppression of the free press had destroyed the entitlements through which laborers and rural workers could convert their labor or income into food. The insight extends with uncomfortable precision to the AI transition.
Sen's empirical work on Bengal overturned the conventional explanation of famine as a Malthusian shortage. The food existed. What failed was the distribution system — the complex of wages, prices, property rights, governmental intervention, and informational flows that, under normal conditions, translated economic activity into nourishment. When the distribution system collapsed, people starved surrounded by food.
The concept operates at multiple scales. At the individual level, an entitlement is the legitimate claim a person has on a resource through existing institutional structures — the wage that buys food, the subsidy that secures medicine, the legal right that protects property. When entitlements fail — through inflation, through institutional breakdown, through the withdrawal of governmental support — the resource may remain abundant while access to it collapses. At the systemic level, the pattern of entitlements within a society determines who benefits from existing resources and who does not.
Applied to AI, the entitlement approach reveals what access-based framings miss. AI capability in 2026 is not scarce. The frontier models are available through subscription. The tools exist. What is scarce is the constellation of conditions — financial infrastructure to pay the subscription, electricity to run the computer, connectivity to reach the service, educational preparation to direct it productively, legal protection for what is built, market access to reach customers — that convert the available resource into capability. Each missing entitlement is a break in the chain between the resource and the person.
The structural parallel to Bengal is precise even though the moral stakes are different. No one is starving from AI deprivation. But the same logic operates: the question is not whether the resource exists but whether institutional structures translate resource availability into human benefit. Sen's crucial finding about Bengal was that famines can be prevented not by increasing the resource supply but by restructuring entitlements — by building the institutional mechanisms that ensure distribution. The same prescription applies to AI: the expansion of human freedom the technology makes possible depends on building the entitlement structures that ensure broad access, and the building is institutional work the technology itself does not perform.
Sen developed the entitlement approach in Poverty and Famines (1981), drawing on his childhood experience of the Bengal famine and decades of subsequent empirical work on famines across multiple countries.
Famine is institutional, not Malthusian. Most famines occur amid adequate food supply, caused by entitlement failure rather than scarcity.
Entitlements are the mediating structure. The institutional mechanisms that translate resources into access for specific people.
Democracy prevents famine. Functioning democracies with free press have never experienced significant famines, because information and accountability force response.
The AI parallel. Capability abundance coexists with entitlement failures that prevent many from benefiting.