Research

Working Papers:

“A Note on Income Effects in Health Care Demand” 

How can income effects help us explain the socioeconomic gradient in health care spending? The key insight is that individuals are making decisions by trading off their marginal costs of forgoing consumption against the marginal benefits of care, and the marginal utility of consumption is a crucial component of this tradeoff. Income effects can explain why the poor underconsume care and why behavioral responses to insurance vary for rich versus poor individuals. Using data from the RAND Health Insurance experiment, I show that low-income individuals do not change their health care consumption in response to subsidies.

“Health Insurance for Redistribution” " (joint with Myles Wagner and Anthony Yu)

This paper considers public health insurance design in an enriched optimal tax framework, incorporating the effects of individual health on utility and labor supply decisions. While optimal tax policy involves trading off the welfare benefits of redistributing to low income individuals against the distortionary costs of taxation, health insurance policy involves trading off the welfare benefits redistributing to poor health individuals against the costs of moral hazard. We quantify these trade-offs in the US Medicaid setting and estimate the optimal policy. Given that existing policies provide more generous health insurance to low income individuals, we quantify the costs of moral hazard separately for the rich and poor by revisiting data from the RAND health insurance experiment. We present new causal evidence that moral hazard costs are small for low income individuals. Combining these estimates with a specification of social preferences and data on the empirical joint distribution of health, medical spending, and income, we recover the optimal policy. A planner who places a high weight on the sick, relative to the healthy, chooses an insurance policy that looks like Medicaid: the optimal health care safety net eligibility threshold is 130% of the Federal Poverty Line, subsidizing 100% of medical spending for low income individuals and 70% for the rest. 

"Contracting Solutions with Ethical Professional Norms" 

I study optimal health care contracting in a principal-agent framework with altruistic providers. I find that financial incentives alone cannot correct for the inefficiencies in the health care system, but financial incentives combined with ethical professional norms (altruism) can. Despite asymmetric information and imperfect agency, systems that pay using global budgets (such as the U.K.) can achieve first best outcomes, but this requires that providers value patient benefits from costlier treatments. Failing to account for altruism and using cost-reimbursement contracts creates incentives to provide low-value care, and high-volumes of these services accrue to significant economic magnitudes.

"Evaluating the Optimality of Provider Payment Contracts" 

This paper proposes a method for evaluating the optimality of provider reimbursement contracts that are partially retrospective, meaning that they condition payment on ex-post reported costs. In a setting where patients heterogeneously benefit from medical care, I derive the optimal linear reimbursement contract for an insurer maximizing the aggregate health of his patients. Then, I propose two ways to empirically calibrate the optimality condition of the insurer with respect to the component of the contract that reimburses retrospectively. I apply one method to the Medicare Outpatient Prospective Payment System, and estimate that 14% of Medicare payments are sub-optimally prospective. 

Research in Progress: 

"Redesigning Payment Policy for Physician Administered Drugs" (with Keith Ericson and Amanda Starc) NIHCM Grant Award

In this paper, we estimate the effects of Medicare Part B payment policy on the price path of physician-administered drugs by comparing drugs where Medicare Part B is a larger versus smaller share of demand. We find that drugs with high Medicare market shares have slower price growth than less exposed drugs, but higher launch prices. We then estimate a dynamic demand model for physician-administered drugs using Medicare Part B pricing and utilization data. Current payment policy (lagged cost plus a percentage market up) for Medicare Part B drugs creates financial incentives for physicians to choose more expensive drugs. We use our demand estimates to quantify the effects of counter-factual payment policies on prescriptions, prices, and the Medicare budget. 

"Optimal Design of Pharmaceutical Contracts for Static and Dynamic Efficiency" (with Keith Ericson and Amanda Starc) 

In pharmaceutical markets, the presence of insurance affects the standard innovation policy trade-off between dynamic efficiency (setting higher prices to provide incentives for innovation) and static efficiency (driving price towards near marginal cost of production). This paper characterizes the effects of patient cost-sharing on investment and pricing incentives. Cost-sharing always augments static inefficiency due to moral hazard, but we can implement first-best dynamic efficiency. Cost-sharing can induce innovation for drugs whose benefits are uncertain at the development stage by aligning market returns with social value, such as anti-psychotics. For drugs whose benefits clearly target a patient population or disease category (e.g. anti-inflammatory) cost-sharing accelerates investment, but distorts the static price far above the monopoly price.  

“A Prescription for Manipulation” (with Sayeh Nikpay and Rena Conti) 

The 340B program entitles certain public, non-profit hospitals to discounted outpatient drugs, and substantial profit margins when discounts are not passed onto payers. Previous work suggests 340B increases ex post strategic behavior. Using hospital administrative data from 1996-2016, we show that 340B also results in ex ante strategic behavior through manipulation of eligibility criteria. Non-parametric density tests show marked discontinuities in the eligibility criterion around the eligibility cutoff. The non-parametric hazard of manipulating eligibility for 340B is higher among hospitals with larger returns to participation. Finally, within-hospital trends in the 340B eligibility criterion break after hospitals meet the minimum eligibility criteria.

“Better and Cheaper Health Care: Inculcating Ethical Virtues in Physicians”(joint with John Rhee M.D., Amish Acharya, and Randall Ellis)

We propose a large-scale randomized controlled trial to test a novel and holistic health care financing model that combines behavioral interventions with financial incentives to achieve cheaper and better health care. The financing model is derived from economic theory, and it combines standard cost-reduction incentives with a behavioral intervention designed to increase physicians’ ethical considerations. Theory predicts that ethically motivated physicians will have a higher likelihood of making care decisions that maximize the long run health of patients, and consequently minimize long run health care costs. We propose testing the theory in the U.K. General Practitioners (GPs) setting. We will randomly assign 200 practices of GPs in the northwest London area to an ‘ethical virtues’ intervention or control. We expect providers in the ethics treatment arm to have a higher propensity to undertake non-reimbursed testing and preventative care procedures that translate to lower lifetime patient health care costs within on three of the most expensive chronic disease areas: peripheral artery disease, chronic kidney disease, and coronary heart disease. Using the data customarily collected by the Primary Care Clinical Research Network, we will quantify the long-run savings of the ethics treatment by tracking patients and comparing their respective costs in the primary and secondary care settings.