Acceso usuarios
Kaname Miyagishima
In this paper, we study an education-planning problem by using a mechanism design approach. We consider a model where agents have different abilities in acquiring education and belong to different social groups (for instance, races or genders). Under the information constraint that the abilities of agents are unobservable but group memberships are observable, we derive two sets of education policies derived under Rawlsian and utilitarian social welfare functions. Our main results show that the utilitarian planner does not discriminate agents by their social group membership, while the Rawlsian planner provides a form of affirmative action policy. We also study second-best optimal education policies in the case of general degrees of inequality aversion. In this case, it is shown that affirmative action is not necessarily supported, and the differences in education levels are determined by the agents’ marginal effects of utility to their group’s aggregated welfare.
