Moral Hazard under Liquidity Constraints

Abstract

Spending induced by health insurance is often called moral hazard and definitionally assumed to be inefficient. We adapt standard models and show that for those living ``hand-to-mouth", the financing benefits of insurance cause a portion of moral hazard to be efficient. Although insurance's price distortions also create some inefficient spending, the net welfare impacts of moral hazard can be positive. We present an intuitive graphical framework and formal results to distinguish moral hazard’s efficient and inefficient components. Simulations show economically significant net benefits of moral hazard in many cases. Our framework also provides a new way of modeling the ``income effect'' induced by insurance, and distinguishes it from the ``liquidity effect". While both can lead to efficient moral hazard, moral-hazard benefits from the ``liquidity effect" are often substantially larger. We use our framework to revisit prior estimates of Medicaid’s value from the Oregon Health Insurance Experiment. For individuals with minimal liquidity, Medicaid's value is more than twice prior estimates. (Joint with Johannes G. Jaspersen, and Justin Sydnor.)