Ulric B. and Evelyn L. Bray Seminar in Political Economy

This paper develops a theory of the effectiveness of government programs. In the model, a bureaucrat chooses a mechanism for assigning a good (such as a license, or a benefit payment) to a client of uncertain qualifications. The mechanism uses a means test to verify the client's eligibility. A politician who values the payoffs of a subset of client types exercises oversight by designating the resources available for means testing and the population of clients that can be served. The model makes predictions about common administrative pathologies, including inefficient and politicized distribution of resources, program errors, and backlogs. When the politician favors marginally qualified clients, programs will tend to be universal, with low per capita spending and high error rates. When the politician favors highly qualified clients, programs have smaller client populations, higher per capita spending, and lower error rates. Notably, the bureaucrat can only use discriminatory testing in the latter case, and programs with larger budgets do not spend more per client.

For the full paper, please click here.