Redistribution with Labor Markets Frictions

How do labor market frictions affect optimal redistributive policies? Are current real world policies efficient? To answer these questions and assess their quantitative relevance we embed an otherwise standard Mirrlees’ (1971) economy in a directed search environment. Effort and employment probability wedges must have the same sign for an allocation to be efficient. Moreover, constrained efficient allocations are shown to be monotonic in both effort and the probability of finding a job, a property that is not implied by incentive compatibility. After showing that labor income tax schedules along with unemployment benefits do not suffice to implement constrained efficient allocations, we describe tax instruments that do. Our optimal policy formulae can be used to derive an allocation which improves upon any constrained inefficient, and cannot be improved upon by local reforms. We calibrate our model to match data for the US economy and quantify the gains from moving to an optimal policy, finding that it leads to a 2.17% increase in Government revenues.