This paper considers tax expenditures from two related perspectives. First, it analyzes how the incentives on Congress to use a tax expenditure change when principal agent problems are considered. For example, it considers whether tax expenditures can reduce moral hazard or adverse selection problems created by delegations to expert agencies. Second, it considers the condition under which tax expenditures should be expected to be redundant with direct expenditures, as many are. The two, principal agent problems and redundancy, are related because redundancy is often seen as a solution to the principal agent problem. The paper concludes that both principal agent concerns and redundancy might lead to an increase in the use of tax expenditures, although the circumstances in which we should expect this are relatively narrow. The paper then examines the example of the low income housing tax credit, concluding that the credit should be replaced with a direct expenditure in the form of increased tenant vouchers.