In our Health Affairs article, we examined the impact on cost-sharing of Health Savings Accounts (HSAs) coupled with high deductible (HD) health plans. We showed that people facing high marginal tax rates who chose HD plans typical of those offered would face reduced cost-sharing and would likely to increase, rather than decrease, their health care expenditures. The policy justification for HSAs has been that they would lead to better control of health care expenditures and our results suggested that this policy justification is frequently invalid.

Last month in the Health Affairs Blog, Uwe Reinhardt correctly points out that our results are not generalizable to all cases. Significantly lower tax rates and/or higher deductibles can generate the conventional result that HSAs generate increased cost-sharing and health expenditure savings. We agree. If HSAs with high deductibles were imposed on the entire population, they would lead (on average) to increased cost-sharing. However, the resulting expenditure reductions would be modest, due to the (desirable) protection of the least healthy people who are responsible for most expenditures. As Reinhardt notes, these plans would disadvantage many lower income and less healthy people.

The individuals most likely to choose HSAs (coupled with HD plans) are those who are, as in our paper, in high marginal tax brackets. We have shown elsewhere that these plans are likely to be unattractive to those in low marginal tax brackets. HSAs are also likely to be most attractive to people who already hold relatively high deductible health plans. Finally, people who switch to HD plans only to take advantage of the new HSA tax savings are likely to prefer the least onerous HD plans available, like those we model in our paper. In this context – the situation that currently exists in the market – our findings show that HSAs may, perversely, have the opposite effect than their proponents intend. Rather than lowering the nation’s health care costs, the availability of HSAs may raise them.

Apparently, proponents of HSAs have used our results of reduced cost-sharing as a marketing device. Such an argument completely turns on its head the original primary reason for HSAs– that they would make patients think more carefully about the costs and benefits of treatments, causing them to cut out certain forms of care or use cheaper alternatives. According to this marketing, the value of HSAs lies not in reduced expenditures but in an expanded tax preference for medical care among higher income individuals. We thank Dr. Reinhardt for pointing out this hypocritical perversion of our findings.