Bureaucracy and Social Transfer Systems
Summary
I construct a model of optimal need-based social transfers with costly bureaucracy. Many real-world transfer programs, such as benefits for people with disabilities, care benefits, or family assistance, aim to allocate money to agents with higher need rather than to redistribute between income groups. The state would like to individualize transfers because agents differ in their marginal utility of money, which is higher for agents with greater need. Bureaucracy makes individualized (case-by-case) transfers possible, but it is costly: for the state, because it has to screen applicants, administer individual cases, and run an increasingly complex system, and for agents, because of the reporting burden. This burden is precisely what allows the state to identify those in need. The state therefore chooses the level of bureaucracy on two margins: which agents receive a uniform transfer and which are treated individually (the extensive margin), and how much burden individually treated agents face (the intensive margin). Even when the state can observe need, it is optimal to give at least some range of agents a uniform transfer, because treating types individually is costly in itself. When need is unobservable, agents with lower need receive an information rent, and the state may impose higher burdens on agents with higher need in order to screen them. Informational constraints can make it optimal to pool agents in the individualized region, either with the uniform-transfer recipients or in additional brackets with higher transfers and higher burdens. I provide interpretable conditions for where to find these pools. For individualized and pooled transfers, I derive formulas akin to the ABC formula of optimal income taxation that separate the roles of need, the type distribution, welfare weights, and bureaucracy costs. My results offer an explanation for the complex multi-bracket structure of many real-world transfer systems.