Consumer perceptions and health insurance decisions

Numerous studies have shown that consumers react imperfectly to changes in health insurance coverage. To justify consumer valuation in health insurance decision-making, I use Medical Expenditure Panel Survey (MEPS) data and conduct three studies to examine consumer’s private information in health insurance decision-making under a conceptual framework of consumer perception, which potentially is informative about Affordable Care Act (ACA) Health Insurance Marketplace consumer behavior. In the first study, I examine the joint role of individual preferences and health risk in two types of insurance decision-making: the probability of being insured and the probability of employment-based insurance if insured. Using logistic regression, I find that the healthier and wealthier consumers tend to have more positive attitudes towards health insurance and thus are more likely to be insured. The effects of health risk measures vary largely in insurance decisions conditional on different preference measures and preference levels. In the second study, I investigate insurance coverage bundle choices with multi-dimensional private information in an artificially created market setting. I adapt the approach developed by Lokshin and Ravallion (2005) and conduct logistic regression modeling to estimate the reduced forms for coverage bundle choice and consumer attitude respectively. Predicted linear indices for consumer attitude and coverage bundle choices are calculated separately, then their correlation coefficients are compared. In this study I find that consumer attitude plays a dominating role in health insurance decision-making, suggesting that risk preferences may internalize health risks and influence insurance purchasing decisions. To further explore consumer perceptions within an individual’s personal system of decision rules, in the third study, I construct coverage bundle choices in an order from the least complete to the most complete, and examine the effect of consumer perceived plan quality to coverage bundle choice decisions. I use the generalized ordered logit method and a Bayesian learning process for the analysis. I find that coverage bundle choice decisions are value-based, for which perceived plan quality plays a significant and persistent role. The study results also have important policy implications to enhancing consumer engagement and optimizing health insurance management to provide high quality care to health insurance beneficiaries