WORKING PAPERS

Dynamic Moral Hazard in Nonlinear Health Insurance Contracts [SSRN]

(Best Paper Award at the IAAE Conference 2021, Honorable Mention) 

Standard health insurance contracts generate nonlinear pricing through the presence of deductibles and caps on out-of-pocket spending, in which the out-of-pocket price paid by consumers decreases as the cumulative use of health care increases. This nonlinear benefit structure, coupled with the uncertainty intrinsic to future health care demand, provides dynamic incentives for consumers’ choices: health care utilization today reduces future expected prices. Standard analyses of insurance contracts study the trade-off between the welfare gains from risk protection and the welfare losses from moral hazard, which I relabel as static moral hazard. In this paper, I study a new source of moral hazard, dynamic moral hazard, which I define as the additional health care utilization when individuals internalize that current utilization lowers future expected prices via the nonlinearities of the contract. By leveraging the random assignment of families to health plans from the RAND Health Insurance Experiment, I am able to focus specifically on moral hazard, avoiding the typically confounding adverse selection present in insurance markets. I develop and estimate a dynamic, stochastic model of weekly health care utilization at the family level that incorporates the dynamic pricing effects. My estimation framework allows for flexibly-correlated multidimensional unobserved heterogeneity related to family health risk, preferences for visiting a doctor, and price sensitivity. I document that 40 percent of total moral hazard is attributed to dynamic moral hazard. Using my model and estimates, I study the welfare implications of dynamic moral hazard in the setting of employer-sponsored health insurance. My results show that the presence of dynamic moral hazard can severely dampen the welfare gains associated with higher cost-sharing and plays a crucial role, distinct from static moral hazard, in determining optimal insurance contract design.

The Dynamics of Preventive versus Curative Care 

Although historically health insurance plans covered preventive and curative care on equal terms, there has been a trend toward increasing coverage for preventive care over the last decade. Additional preventive care decreases future curative care costs by lowering the probability of developing an illness and/or decreasing its severity. In this paper, I study the implications of dental insurance design for both preventive and curative dental care demand and the consequences for oral health. I focus on dental care because it offers a clean environment to distinguish preventive versus curative treatments and their differential impact on oral health. Using detailed dental claim data grouped into episodes of care from the RAND Health Insurance Experiment, I build and estimate a dynamic dental health capital model where individuals can consume three types of dental care: primary preventive care, which prevents the development of oral disease (e.g., cleanings); secondary preventive care, which include activities taking place before disease is recognized (e.g., examinations); and curative care, which includes the treatment of diseases (e.g., fillings). The model incorporates both the possibility that preventive care triggers a short-run surge in curative care demand through the detection of oral diseases and, simultaneously, a long-run reduction in future dental care costs through endogenous oral disease development and severity. I also study how nonlinear dental insurance contracts affect the relative level of preventive and curative dental care demand and subsequent oral health outcomes. 

Cash Transfers, Maternal Health, and Children's Birth Outcomes, with Timothy G. Conley

Timely prenatal care and maternal health are important determinants of birth and child outcomes. Many developing countries implement conditional cash transfer (CCT) programs to unemployed pregnant women, which typically condition the transfer on obtaining prenatal care and participating in meetings about proper nutrition. Existing literature on the evaluation of maternity CCTs finds contradicting results. In this paper, we study the impact of the “Asignación Universal por Embarazo” (AUE), a maternity CCT program in Argentina, on maternal health and children's birth outcomes, and investigate the reasons for the low take-up rates. We build a unique dataset that links multiple sources of hospital records with administrative micro-data spanning seven years regarding individual women's pregnancy, timing of checkups, and birth outcomes. To isolate the story of lack of information about the program from other explanations for low participation rates such as stigma, we exploit the difference in launching time of the AUE and an already established CCT program implemented by the same government agency that targeted unemployed women with children under the age of 18. Specifically, eligible women who receive this latter program face significantly lower information costs to apply for our program of interest.

Unveiling the Energy Price Elasticity: Exploring the Impact of Price Shocks through Regression Discontinuity Design, with Alejandro M. Danon, Jared Gars, Mariana Kestelman Borges, and Lourdes Zulli

Understanding how electricity demand responds to price shocks is a key question for a number of actors along the electricity supply chain as well as policy makers, albeit its estimation present several challenges. In this paper, we exploit a natural experiment to estimate the short-run impact of a price shock on residential electricity consumption. In particular, in January of 2021 the utility company adopted a new tariff schedule whereby the fixed component of the tariff was organized in four tiers based on households’ annual moving average consumption, which we exploit in a regression-discontinuity design. Despite the large average price increases at each fixed-cost cutoff, we find no significant effect of the tariff change on subsequent electricity consumption around the three thresholds. This lack of demand response to prices suggests that non-price instruments may be more effective at influencing residential electricity consumption.

WORK IN PROGRESS

The Role of Health Insurance Generosity and Doctor Prescription Behavior on Opioid Painkiller Use and Addiction, with M. Antonella Mancino

(Supported by SSHRC Insight Development Grant - Co PI)

Pricing Pain: The Behavioral and Fiscal Impact of Opioid Taxes, with M. Antonella Mancino

(Supported by SSHRC Insight Development Grant - Co PI)

Poverty Penalty: The Harm of Mismatch for Low-Income Households, with Alejandro M. Danon and Fernanda Marquez Ragonesi

Climate Change, Seed Insecurity, and Farmer Social Networks in Malawi, with Timothy G. Conley and Rowena Cornelius

- Fieldwork completed -