MSA’s — Panacea or Placebo?

The policy debate over whether to extend tax benefits to Medical Savings Accounts (MSAs) has tended to polarize into two camps. MSA advocates gravitate toward sales hype and Panglossian promises while opponents hurl back knee-jerk claims that widespread adoption of MSAs will create windfall gains, revenue losses, and fractured insurance risk pools.

The respective hopes and fears of MSA partisans and critics have collided during last year's attempt to restructure the Medicare program and, more recently, in the midst of the current effort to adopt incremental "reforms" of health insurance markets. Efforts to reach a conference agreement between House and Senate health bills adopted this spring have remained fundamentally stalemated for several months over the MSA issue. A recent analysis of MSAs by RAND researchers provides new scholarly light amidst the partisan heat on Capitol Hill. It suggests that MSAs offer neither a surefire cure nor a dangerous curse for the complex problems of health insurance markets. The strengths and weaknesses of MSA proposals depend upon the details of their design.

RAND concludes that MSAs would have little impact on the health care costs of Americans with employer-provided insurance ("the MSA approach is not likely to produce the reduction in health care use that its advocates foresee"), but it knocks down most concerns about adverse selection. Depending on the size of the stop-loss limit on out-of-pocket spending provided under a catastrophic insurance policy, RAND finds that "MSAs would be attractive to both sick and healthy people" and, if health plan choice is taken into account, enactment of federal MSA legislation would change total health spending by between -2 percent and +1 percent.

The RAND analysis is particularly noteworthy and credible because the primary scholarly foundation and impetus for MSA proposals come from the oft-quoted RAND Health Insurance Experiment, which examined the effects of various cost-sharing provisions on health insurance expenditures from 1975 to 1982. Although MSA advocates have frequently cited the RAND experiment as evidence that the combination of MSAs and high-deductible, catastrophic insurance plans will reduce total health care spending, they invariably neglect to adjust for the fact that the out-of-pocket, cost-sharing component of health risk under the RAND experiment received no tax advantages, but MSA funds do in current proposed legislation. Depending on the manner in which cash contributions to MSAs are treated as tax-exempt, such tax advantages may stimulate greater health care spending to varying degrees. MSAs whose annual cash contribution amounts to a substantial portion of the deductible under the accompanying insurance plan will, in effect, reduce the "after-tax" coinsurance rate on both uncovered health care services and other out-of-pocket (below the deductible) spending on health care. That will generate more spending on those items. (The subsidy could vary from 15 percent to as much as 50 percent, depending on a taxpayer's marginal tax rate bracket and whether the MSAs are funded by employers or employees.)

Moreover, as noted in a 1995 American Academy of Actuaries study of MSAs, accumulation of cash balances in an individual's MSA can counteract the health care utilization reduction resulting from high copayments under an accompanying catastrophic insurance plan. The greater the amount contributed to the MSA, the lower the "real" stop-loss cap (or deductible) on annual out-of-pocket spending. The RAND analysts observe that, after exceeding their annual cap, insured families face no cost sharing for the rest of the year and therefore will consume more health care services. The Actuaries' study further frames the cost containment issue for MSAs in terms of whether the insured party views his MSA balance as simply another form of insurance (leading to higher health care utilization) or as personal savings (reducing utilization). The RAND analysis similarly asks whether the funds in MSAs are valued as equal to fully fungible money. Although their primary finding assumes 100-percent fungibility, they note that if people value MSA funds at only 75 percent of money that is fully liquid, total health spending increases by 2 percent.

Under most legislative vehicles for tax-advantaged MSAs, proposed federal tax treatment will encourage more, rather than less, health spending on the margin. Partly for myopic budget-scoring reasons, withdrawals from MSA accounts for non-health-care spending generally are subject to both income and payroll taxes (previously deferred), as well as a 10-percent penalty before the age of 59 1/2. As a result, spending an additional dollar out of MSA money on health care items would go further than spending that dollar for other purposes.

Thus, the core features of current MSA proposals partly work at cross-purposes to their purported goal of health care cost containment. Further dilution of the spending-dampening effect of the high- deductible insurance plans paired with MSAs can come from problems in benefits coordination for families with additional health coverage through another employer (e.g., dual-earner couples can transfer out-of-pocket risk to one employer while gaining increased MSA fund accumulations from the other). It also remains an open question as to whether an employer, in response to a tax-advantaged MSA option, will contribute a higher percentage of overall employee compensation to the health care benefits package (to leverage the tax benefits for out-of-pocket health spending) or keep most of its savings on health insurance premiums for the company and pay the rest in either higher taxable wages or MSA deposits (see, e.g., the initial year of the "bonus" incentive plan by Forbes magazine).

Indeed, one could make the case that political advocacy of tax-advantaged MSAs has increasingly shifted toward emphasizing their potential to enhance and augment health care benefits, rather than reduce overall spending levels. Antipathy to the constraints (real and perceived) of managed care by both patients and physicians has provided an emotionally appealing message and powerful mobilizing force for MSA proponents. MSAs promise the preservation, if not the expansion, of less-restrictive and hassle-free insurance coverage of fee-for-service-style medicine. Of course, as Brookings Institution health analyst Joe White notes, if health benefits are higher under MSA-style insurance plans, the claim of accompanying savings on overall health care costs seems contradictory.

Although MSA proponents draw upon growing consumer disenchantment with the excesses of managed care for popular support, the many large- and medium-sized employers that have reduced their health benefits expenses through various degrees of utilization controls, integrated care, selective contracting, gatekeeper incentives, and case management remain more ambivalent at best. (Indeed, even the primary proponent of MSAs, the National Center for Policy Analysis, has integrated selective contracting and network incentive elements of managed care into the MSA-based health plan for its own employees, in order to influence their choice of physicians.)

In a 1994 review of the potential of MSAs, health policy analyst Gail Jensen found that health care expenses between the "catastrophic" insurance plan and the HMO alternative plan in the RAND experiment were statistically indistinguishable, and she concluded that MSAs "are about equally effective as HMOs at containing costs." However, the RAND analysts recently noted that many current HMOs are more assertive about rationing care than was assumed in their original experiment, and therefore shifts from less-costly HMOs to MSAs would result in less cost savings than their model predicts above. Of course, the qualitative preference of many consumers for greater choice of their physicians and control regarding their treatment plans could well override marginal quantitative differences in health care costs. What, then, is one to make of claims by MSA advocates that thousands of employers have already reduced their overall health benefits costs by 10 to 30 percent, through adopting combinations of high-deductible catastrophic insurance coverage and generously funded MSAs for their employees? First, congratulate them for overcoming the burdens of the limited bargaining clout, undiversified risk profiles, and scarce benefits administration resources that are prevalent in small-group markets and tend to result in overpriced and overinsured health care coverage.

Second, note that these MSA success stories occurred without any new federal tax advantages. Because the MSAs were funded on a post-tax basis, the employers' successes in controlling health benefit costs essentially prove that high-deductible insurance is less expensive and reduces overall health spending more than low-deductible insurance coverage does. No surprise there. Employees in such "MSA/Catastrophic Insurance" plans actually had a portion of their total compensation shifted from health insurance premium payments to additional "wages" (MSA contributions) that were subject to income taxes.

Third, RAND analysts also note that the company experiments with MSAs had no control groups, that firms and employees currently using non-tax-advantaged MSAs are a select group, and that many other firms in recent years have had similar success in curtailing health premium increases through other non-MSA means.

Fourth, many illustrative comparisons between out-of-pocket exposure in lower deductible/comprehensive insurance plans and high-deductible/catastrophic insurance plans both overstate the effect of additional coinsurance and understate the effect of stop-loss limits under the former. Unlike Medicare fee-for-service coverage, the maximum out-of-pocket exposure under most private, low-deductible, comprehensive insurance is not much greater than that promised under MSA/catastrophic insurance combination plans.

Fifth, most professional actuarial opinion remains skeptical about claims that premium savings from high-deductible insurance policies are large enough to provide enough funds in the accompanying MSA to cover most of the out-of-pocket risk under the higher deductible for the general population. The standard actuarial assumption remains that a one-percent increase in out-of-pocket health spending results in approximately a 0.2 percent reduction in total medical spending. If that relationship holds in most health insurance markets, then the dollar reduction in health benefit payments and total expenditures (by switching from a comprehensive insurance plan to a high-deductible plan) must remain much less than the increase in deductible levels. For example, the American Academy of Actuaries estimated that, for an increase in the deductible from $200 to $1,500 while an employer held health benefit expenditures constant, the premiums for the health plan would decline to a level ranging from $468 to $552 (with that balance deposited in an MSA).

Although the cost containment case for MSAs is cloudy, charges that their widespread adoption will aggravate risk segmentation and worsen insurance coverage for the sick in the below-age-65 market are overstated. RAND analysts effectively point out that MSA plans can benefit both the healthy and the very sick (the latter because of relatively low stop-loss caps, greater choice of physicians and treatment plans, and MSA fund balances available for first-dollar coverage). Benefits for individuals with high medical expenses will be further enhanced if MSA legislation allows them the option of funding their tax-advantaged MSAs themselves (rather than relying on employer contributions).

Distributional analysis of "winners and losers" under MSA-style coverage overlooks the real benefits of expanding choice and competition in health insurance markets. The skewed distribution of health care claims in any given year means that, under a broad-based, comprehensive, low-deductible insurance plan, most insureds are net "losers" on a cash payout basis in any single year. The flip side of reducing the scope of third-party insurance coverage for early-dollar claims and in effect "refunding" the cash-equivalent of foregone insurance benefits to insured individuals is a partial elimination of the previous insurance cross-subsidies from the healthy to the sick. Because most insured individuals in any given year will incur modest, if any, health care charges, they will "win" under a MSA plan that returns cash to them.

However, the point of true insurance is to provide protection against unanticipated risks that an individual cannot finance on his own. The broader trend of reduced cross-subsidization of health insurance policies by employers reflects both the increased price competitiveness of commercial insurance markets and employers' desires to make their own workers more aware of the costs of their insurance.

In summary, tax-advantaged MSAs may have a mixed impact at best on health care spending levels, but adverse selection concerns are exaggerated for the employer-provided group insurance market. In the next issue of UpDate, we will examine the political benefits of MSA proposals, as well as how to provide a market-based overhaul of health care reform proposals.

–Tom Miller