When is full health insurance not optimal

In many markets, however, we observe that lower income individuals do not purchase the most generous forms of health insurance. For instance, despite the passage of the Affordable Care Act–which provided subsidies to lower income individuals to purchase relatively generous insurance–some individuals preferred the less generous Association Health Plans (AHPs) also known as “skinny” health plans.

A key question is, our lower income individuals making rational decisions? One could consider individuals to be not rational if they have hyperbolic preferences (i.e., focus on short term cost of premium and ignore long-term health care costs) or over/underestimate illness probabilities.

A recent paper by Jasperson et al. (2022), however, uses an expected utility framework and shows that in many cases it is rational for individuals to purchase cheaper, less generous insurance coverage, especially if individuals have a lower initial endowment of wealth As wealth increases, however, people tend to buy more generous insurance. While the paper itself provides the mathematical proof, the logic is outlined in the excerpt below.

As individuals obtain more wealth, the marginal utility of non-medical consumption decreases. The cost of improving one’s health is unaffected by wealth. Because the marginal utility of health at the very least decreases less than the marginal utility of consumption, spending money on healthcare becomes more attractive…Hence, underinsurance due to restricted substitution between health and wealth is more likely in the low-income population.

The paper provides some logic for why “low-income population often foregoes purchasing healthcare even in the presence of liquidity reserve.” Empirically, we also do see lower income individuals more likely to trade off health for wealth by being more likely to accept a higher paying job with more health risk (e.g., mining).

Note that one of the key assumptions of the paper is that individuals are prudent over consumption (i.e., they preferred right skewed consumption distributions; formally, the third partial derivative of the utility function with respect to consumption is positive).

This content was originally published here.