When eligible people fail to receive a public benefit, the shortfall is often described as an implementation problem. A legislature creates a program; agencies then surround it with long forms, documentation demands, interviews, and repeated renewals; participation falls below the level lawmakers intended. On this account, administrative complexity is friction in a machine. The remedy is to simplify enrollment until take-up more nearly matches legal eligibility.

Research on administrative burden has complicated this picture. Burdens have at least three components: learning costs, such as discovering that a program exists; compliance costs, such as collecting records or traveling to an office; and psychological costs, such as stigma or the loss of autonomy involved in proving need. These costs are not evenly distributed. A salaried worker with internet access may upload a pay stub in minutes, while a worker paid irregularly in cash may spend days reconstructing income that no employer has documented. The same formal requirement can therefore sort applicants by clerical resources rather than by legal entitlement.

Defenders of screening respond that some friction is useful. Benefits are scarce, errors undermine public confidence, and documentation can deter fraudulent claims. In this view, eliminating every hurdle would trade one failure—eligible people excluded—for another—ineligible people paid. The response has force, but it does not establish that existing burdens are calibrated to accuracy. A requirement may persist because it is familiar, because agencies fear criticism for visible overpayments more than for invisible nonparticipation, or because elected officials can reduce enrollment through procedure without openly narrowing eligibility. Burden can be policy by other means.

That possibility changes how reform should be evaluated. Counting the number of form fields is not enough. Analysts must ask which error a requirement prevents, who can satisfy it, and what happens when they cannot. Automatic enrollment from reliable tax records may reduce both exclusion and fraud; by contrast, an in-person interview scheduled only during working hours may add little information while screening out people with inflexible jobs. Nor is a burden justified merely because every applicant formally faces it. A staircase is the same for everyone, yet it does not impose the same cost on everyone.

Administrative procedures are unavoidable, and some verification will remain necessary. But procedures should not be treated as a neutral conduit between legislation and outcome. They help determine who receives what the law appears to promise. The proper comparison is therefore not between a rigorous program and a careless one, but among verification methods with different informational value and distributive effects. Once burden is recognized as a substantive design choice, low take-up can no longer be presumed an accidental defect; it may be the predictable result, and sometimes the concealed objective, of the way a program is administered.