The Effect of Health Savings Accounts on Group Health Insurance Coverage

This paper presents new empirical evidence about the impact of Health Savings Accounts tax subsidies on group insurance coverage. To make individuals save for future health care expenditure, and become price sensitive to medical care cost, the federal government established Health Savings Accounts (HSAs) to provide a federal income tax exemption for the out-of-pocket portion of eligible high-deductible health plans (HDHP). Following with federal law, state laws and regulations took steps to promote the use of HSAs by complying with the federal guidelines and, in some cases, providing a state income tax deduction for HSAs contributions from 2003 to 2011.

This tax program shifts tax subsidies from the premium component in traditional plans to the out-of-pocket component in HDHP, and it may expand the group health insurance market. Previous research has widely discussed the effectiveness of HSAs, but empirical evidence is still lacking. My study adds to literature in two ways. First, by using state by time variation in income tax subsidy policy for HSAs and nationally representative individual level data, I am able to identify the effect of HSAs subsidies on group insurance coverage directly, not only for all workers, but also workers from different sizes of firm and age groups. Second, to further examine the effect of the subsidies’ magnitude, I translate the estimated effect of HSAs subsidies into a parameterized estimate of the relationship between the state tax price of HSAs and group insurance coverage. The top federal rate together with maximum income tax rate by state and year are imputed to calculate the HSAs state tax prices.

The analysis in this paper is based primarily on a repeated cross-section data from the 2004 to 2012 March Current Population Surveys (CPS). State law information is mostly from National Conference of State Legislatures. What is more, the top marginal federal and state income tax rates are calculated with the TAXSIM developed by NBER. Conclusions are as follows: I find no statistically significant effect of HSAs income tax subsidies on whole group health insurance coverage. On the other hand, these subsidies increase small group coverage by a statistically-significant 3 percentage points. I also find that for the people who may make catch-up contributions to HSAs (55-64 years old), HSAs subsidies was associated with around 2 percentage point increase of group insurance coverage. 

Furthermore, the marginal estimated effect of HSAs state tax price on the probability of being covered by group health insurance is negative and insignificant for both whole employees and older workers. But the marginal effect of interaction term of HSAs state tax price and small firm on small group insurance coverage is -0.189, at 5% significance level. It indicates that when the tax price of HSAs decreases 10 cents, the small group insurance coverage will increase by almost 2 percentage points. These conclusions are consistent cross several specifications of the effect of HSAs subsidies, pass a pre-trend check, and are robust to the inclusion of state and year effect.