Health Savings Accounts: will they impact markets?

INTRODUCTION

In the over 30 years since the passage of the Employee Retirement Income Security Act of 1974 (ERISA), retiree income security risk burdens have shifted steadily from employers to households. A primary avenue for this redistribution of risk burdens has been the transition in employer pension plans from a defined benefit (DB) to a defined contribution (DC) structure. As a result, households now bear most of the risks associated with retirement income security. (1) Recently, employee and retiree health insurance programs began undergoing a similar change, with a growing number of current and future retirees expected to bear a larger burden of retirement health care risk. In this paper, we examine the effect of Health Saving Accounts (HSA) on the distribution of relative risk burdens and the price sensitivity of health care purchases. We find that HSAs shift some health care risk onto households and government but are not likely to be very effective at reducing health care demand.

Over the past ten years, health care risk burdens have progressively shifted onto households. Recently, federal tax policy began subsidizing this transition through tax--exempt Health Savings Accounts (HSAs), which require high--deductible (sometimes referred to as catastrophic) health insurance contracts. One goal of HSAs is to increase incentives for "consumer driven" health care demand. That is, the goal is to make households more price sensitive to health care purchases by increasing out-of-pocket spending. Another goal of HSAs is to facilitate an expansion of group health insurance because employers, and in particular small employers, face relatively less risk to future compensation costs when compared to more traditional group contracts with higher premiums and first dollar coverage. Both goals are facilitated through changing the emphasis on health insurance tax subsidies from the premium component to the deductible component of a health insurance contract.

We find that an HSA contract changes relative prices and increases risk burdens faced by households. Both changes have an ambiguous effect on health care demand because the change in health care tax subsidies may create incentives that substitute one set of market imperfections for another. First, we note that HSA-eligible policies redistribute risk burdens away from insurance pools and towards household self-insurance. This will tend to decrease health care demand. Second, HSAs slightly increase expected variable cost but can notably decrease the marginal cost of the deductible component of household health care spending. By placing more emphasis on self insurance (through the deductible), HSAs can reduce the moral hazard problem of over-consumption of health care related to a lack of price sensitivity. (2) Moreover, the potential reduction in the marginal cost of health care may actually increase household health care demand. The net effect on the market from all these changes is ambiguous.

For firms with employer--sponsored insurance, HSAs may exacerbate adverse selection problems, especially if there is a range of insurance benefits offered. For firms with multiple offerings, we expect healthy types to adopt HSA--eligible contracts more readily, while less-healthy types persist in their proclivity for traditional contracts. Given that there are substantial overhead costs associated with any insurance contract, (3) the ability of higher deductibles to reduce premiums is limited, and healthy types may opt out of the group insurance market if premiums do not fall proportionately with the increase in deductibles, leaving the potential for an adverse selection spiral. This suggests the possibility of a separating labor market equilibrium, where older workers experience job lock to compensate for unanticipated wealth losses due to loss of pooled health insurance assets, marginally disabled workers seek full disabled status to secure access to Medicare and Medicaid, and younger or healthier workers facing lower search costs experience an increase in labor mobility and self employment. When so many incentives change by the same policy, it is unclear what the total impact of the policy will be on health insurance and labor markets. The ultimate impact of HSAs will depend on the observed magnitude of future expansions of group health insurance offerings by employers, patterns of employee take-up, and increases in household price sensitivity to health care.

The remainder of the paper is structured as follows. The second section examines recent trends in health insurance pricing and health care consumption. The third section analyzes the impact of various tax subsidies of the relative distribution of health care risk burdens. The fourth section provides a discussion of potential effects on employment-based health insurance. The fifth section offers concluding thoughts.

RECENT TRENDS IN HEALTH INSURANCE

Figure 1 provides information on recent trends in the private health care market and demonstrates what many feel are unsustainable growth rates for health care spending and health insurance premiums. Panel A describes relative changes in premiums and the general consumer price index over the past decade. Panel B describes increases in per-capita medical spending and per-capita GDP.

[FIGURE 1A OMITTED]

[FIGURE 1B OMITTED]

Both panels point to a growing share of income devoted to health care consumption. In addition, premiums have recently grown at rates in excess of ten percent per year. These increases have placed stress on the employer-provided health insurance market, and may be a factor in the observed low growth of real cash wages because employers will direct large proportions of the increase in total compensation into rapidly growing health insurance premiums.

Given that there are real substitution and income effects associated with rapidly rising premiums, households may be better off with insurance contracts that reduce the average fixed cost of health care. In this sense, HSAs might be effective in reducing the tax distortions associated with traditional health insurance by reducing incentives to purchase first dollar care.

There are several reasons to believe that HSAs will have a limited effect on premiums. First, Figure 1 suggests that changes in premium growth rates tend to lag changes in the growth of health care spending. This is mainly because actuarial methods tend to rely on prior risk experience when accessing future premiums. Even then, the impact of HSAs will have to be disentangled from other market changes--changes in health care utilization due to changes in morbidity, longevity, and medical technology, for example. Thereby, changes to the relative composition of the three-part pricing form of premiums, deductibles, and co-insurance that are associated with the introduction of HSAs may not occur for several years, and may be subsumed by more significant changes in the health care market. To the extent that HSAs do not change the final equilibrium composition of contract components, they may still be seen as a partial remedy--subsidizing household saving to meet the increased burden associated with market trends away from pooled insurance. What is not clear is how HSAs affect health care demand, since HSA tax subsidies lower the marginal cost of health care for households because HSA balances are used to finance the deductible and co-insurance components of health care. Given that health care is a normal good, HSAs may increase consumption since the government is subsidizing the marginal dollar of care. Also, unless preventative health care is covered by the pooled insurance component, the fungability of HSA tax advantaged savings provides an incentive to under-consume cost-effective health care products and increases the probability of later catastrophic claims on the pool Thus, either preventative care must be included in pooled coverage or, in lieu of coverage, future catastrophic claims should be expected to increase demands on the pool. For all of these reasons, the opportunity for high-deductible HSA-eligible policies to stem premium growth should be seen as limited.

A focus on traditional public insurance is motivated by recent trends. Indeed, while the percentage of uninsured Americans has not changed all that significantly in recent years, in 2004 the percent of workers with employer health insurance dipped below 60 percent (DeNavas-Walt, Proctor, and Lee, 2005). Public programs are increasingly used as the primary health insurance providers for working-age adults as well as for more traditional populations of the very young (through the State Children's Health Insurance Program--SCHIP) and the aged (Medicare). While Medicare is often held up as an example of an efficient health insurance program, as Figure 2 shows, Medicare costs and spending have risen rapidly as well.