The paper is divided in two parts: the first presents a simple static model of education choice, labor supply and consumption. Here the only source of uncertainty is ability, from the perspective of someone before they are in a position to make any choice (say at birth). The simplified static model is used to explore results analytically, comparing the first best planner’s solution, the outcome in competitive equilibrium (CE) with no policy intervention, and the decentralized solution with optimal policy. The authors limit themselves to a narrow set of policy instruments: an education subsidy for college attendance, an amount of income that is deductible for tax purposes and a flat tax rate applicable to all remaining income. With this simple set of policy instruments they explore the role of progressivity as an insurance mechanism. Within this static model they show that the competitive equilibrium will deliver efficient levels of labor supply but that consumption allocation will differ from the solution of the planner׳s problem, providing too little insurance. The efficiency result is driven by the quasi-linear structure of the utility function. If this were not the case a distortion in the level of consumption would also lead to a distortion away from efficient labor supply, through income effects or for richer types of preferences through the possible complementarity between hours of work and consumption. It is possible to provide some level of insurance with optimal choices over the available policy instruments, but these instruments alone are not sufficient to implement the planner׳s solution in a competitive equilibrium. To implement the planner’s solution requires taxes that can depend on ability, which they assume is not possible. Thus, a system of education subsidies, progressive taxation and lump sum transfers is second best.