Incentives and the sorting of altruistic agents into street-level bureaucracies

Many street-level bureaucrats (such as caseworkers) have the dual task of helping some clients, while sanctioning others. We develop a model of such a street-level bureaucracy and study the implications of its personnel policy on the self-selection and allocation decisions of agents who differ in altruism towards clients. When bureaucrats are paid flat wages, they do not sanction, and the most altruistic types sort into bureaucracy. Pay-for-performance induces some bureaucrats to sanction, but necessitates an increase in expected wage compensation, which can result in sorting from both the top and bottom of the altruism distribution. We also show how client composition affects sorting and why street-level bureaucrats often experience an overload of clients.