This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

C O M M E N T A R Y
U N I V E R S A L  C O V E R A G E
W E B  E X C L U S I V E   

31 March 2004  
  

The Institute Of Medicine Committee’s Clarion Call 
For Universal Coverage

The recommendation for universal coverage faces an uphill struggle,
because of difficulties defining benefits and variations in health care costs.

By Joseph P. Newhouse and Robert D. Reischauer

ABSTRACT:

The Institute of Medicine (IOM) Committee on the Consequences of Being Uninsured is to be commended for its work, which recently culminated in the release of six volumes on the subject. The concluding volume presents a vision for universal coverage and describes four options for achieving it. The options include an incremental approach, employer and individual mandates, and a single-payer plan. We identify complications involving benefits and geographic variation in costs surrounding attempts to achieve universal coverage. The complications suggest that the committee’s cost estimates may be too low and that there may be sizable political barriers to the proposals.


The institute of medicine (IOM) Committee on the Consequences of Being Uninsured is to be commended for its series on the uninsured in America.¹ The six volumes examine who lacks insurance coverage; detail the consequences of uninsurance for individuals, families, and communities; measure the economic costs of being uninsured; and make recommendations for the future. Not surprisingly, the information on the nature and adverse consequences of uninsurance becomes sketchier as one moves from the individual to the family, the community, and the nation. Nevertheless, the reviews of the literature contained in these volumes, by themselves, are worth the price of admission.

The IOM committee, of course, hopes that its labors do more than simply assist future students of the uninsured with a first-class literature review. Indeed, it would be happiest if there were no further need to study this subject. To that end, the concluding report offers a “vision statement” and a set of principles, the first and “most important” of which is that “health care coverage should be universal.” However, when it comes to drawing a roadmap of how to proceed, the subject of the final volume, the committee loses some steam.

The committee puts forward four options or prototypes. The first, an incremental approach, would expand Medicaid and State Children’s Health Insurance Program (SCHIP) eligibility, lower the age of Medicare eligibility to age fifty-five, and provide refundable tax credits for health insurance. Unless the eligibility extensions and tax credits are much more generous than anyone is now proposing, this option is unlikely to come close to achieving universal coverage, a fact that the committee readily concedes. By contrast, the committee’s remaining three options could, in theory, achieve universal coverage. The second option follows the general contours of the Clinton plan, an employer mandate coupled with a public program for those not covered through work. The third option involves individual mandates with tax subsidies, and the fourth, a single-payer, public plan.

The first (incremental) approach has the virtue of being politically and institutionally more acceptable than the other three. Furthermore, in today’s constrained fiscal climate, some sort of incremental approach is by far the most likely course and the one we prefer to no action at all.² Nevertheless, incrementalism is unlikely to get close to what most would consider universal coverage. Because of this and because the committee believes that there should be universal coverage, we focus our comments on two complications that arise when pursuing any of the other three approaches, although the same complications also arise if tax subsidies are used as part of an incremental approach. Specifically, we examine (1) the nature of the benefit package and (2) geographic variation in health care spending. These two issues, which the committee did not address, involve sizable redistribution, which could raise barriers to seeking universal coverage.

The Benefit Package

The IOM committee tried to make the case that universal insurance is not costly relative to what the United States already spends on health care services. It estimated that universal coverage would boost spending, in 2001 dollars, by $34– $69 billion—about 3–6 percent of total spending on personal health care services. These estimates, of course, depend critically on the nature of the benefit package—what services are covered, the cost sharing for covered services, and required premiums—which affect not only the overall costliness of the program but also the distribution of cost among individuals and the amount of services used.³

But even cutting the committee some slack, its estimate is likely to be unrealistically low. The rub comes in the need to work through how, in the real world, the guaranteed and specified benefit package might affect people who are now insured.

While the committee’s estimate assumes that the uninsured would be provided with an average insurance policy, roughly half of those who are already insured have below-average insurance policies. It seems unlikely that policymakers would mandate or provide those who are now uninsured with a more generous policy than the policies covering many of the currently insured. If the subpar policies of the insured were beefed up, the cost of universal coverage would have to include both the increased use of services by the uninsured and the additional service use by the currently (perhaps under-) insured.⁴ Alternatively, the committee’s cost estimate could be achieved by providing the uninsured with a below-average policy and raising those who are currently insured with policies less adequate than this to the new minimum, although that would dilute the gains the committee estimated, since many of the studies the committee used implicitly compared the uninsured with a group of people having an average policy.⁵ In either case, there would be opposition from those—workers, employers, or taxpayers—who were asked to pick up the tab for augmenting substandard policies.

The point here is that the cost of providing the uninsured with an average insurance policy will be greater than simply the cost of the additional services that the uninsured would use if they had such coverage. Moreover, any proposal to provide coverage to the uninsured would meet strong opposition from those who were asked to foot the bill for the unavoidable requirement to bring subpar existing coverage up to the level of that provided to the uninsured.

The issue becomes even more complex in the case of a single payer with a single plan; for example, some advocates have suggested covering everyone through an improved version of Medicare. Not only does the generosity of the single benefit package become an issue at both ends of the distribution, but provider payment policy becomes a battleground. If the single-payer plan has benefits that are around the current average, roughly half of people with above-average coverage now will have less coverage under reform, something they are unlikely to take kindly, especially if their current benefits had been decided through collective bargaining. And if the single plan is much better than the current average, costs explode.

If the single-payer plan were modeled after Medicaid, beneficiaries might be pleased because the plan would cover a wider range of services than most plans now include and would require negligible cost sharing. But providers would be furious because the rates at which Medicaid pays providers are notoriously low. If Medicaid were the single payer, would its rates rise? If so, that would add to government and probably social cost, beyond simply the cost of covering the uninsured. If, however, the plan paid for all services at something approximating Medicaid rates, there would be a sharp reduction in resources flowing to health care providers, something that is not likely to happen.

A single-payer approach, of course, need not imply a single plan. Medicare is a single payer for people age sixty-five and older, but most Medicare beneficiaries have a choice of receiving coverage through the traditional fee-for-service delivery system or through a Medicare managed care plan. But if multiple plans are permitted, decisions must be made about the minimum and maximum benefits that health plans have to provide to qualify for participation.⁶

In fact, the issue of what benefits are to be provided must be addressed under all four of the committee’s options. An employer or individual mandate must define the dimensions of the mandated policy, and a tax subsidy must specify the elements of the policy that qualify for the subsidy. Benefits and cost sharing can be specified, or a minimum actuarial value can be set. Either way, additional costs from improved benefits among the currently insured population are likely.⁷

Geographic Variation

A second issue that must be confronted is the large geographic variation in health spending. For example, health care spending per capita in Massachusetts is almost double that in Wyoming.⁸ Should this variation be reflected in the approach? If so, to what extent? For example, if tax subsidies are part of the plan, will they be larger in Massachusetts than in Wyoming? Will they be open-ended? Will they be capped, as in President Bush’s proposal?⁹

Historically, the individual income tax code has not explicitly reflected geographic differences in spending levels, costs, or cost of living. However, some deductions and credits are open-ended, allowing those living in high-cost or high- spending areas to reap larger tax benefits. The deduction for medical spending above 7.5 percent of adjusted gross income is an example of an open-ended deduction. The home mortgage interest deduction is the most widespread example of a capped deduction.¹⁰ Most other income tax parameters, such as the personal and standard deductions and child credits, are uniform across the United States.

If the subsidy for health insurance were uniform or capped, the increment in insurance coverage that the subsidy would buy would surely be less in Massachusetts than in Wyoming. If, on the other hand, the tax subsidy were to vary geographically, difficult and politically divisive decisions would have to be made to define the areas over which the subsidy would vary and to determine how much variation to allow. States, clearly, are not useful geographic units for the first purpose. Health spending is greater in New York City than it is upstate. Even within metropolitan areas, health spending varies considerably. To determine the appropriate amount of variation to provide in the subsidy, policymakers would have to know the variation across areas in the cost of providing the mandated minimal benefit package, information that is not now available. The committee’s second and third options, which impose mandates on individuals or employers, raise these issues because they call for subsidies for those with lower incomes.

Medicaid and SCHIP resolve the issue of differential subsidies by varying the federal matching rate inversely with the per capita income of the state. But this is a notoriously contentious issue, and there would be more dollars at stake in subsidies to move toward universal coverage. Even in Medicare, geographic variation is a contentious issue, as the recent Medicare legislation showed, with provisions such as lowering the labor share for hospitals (to benefit rural areas) and updating health plan rates at 100 percent of fee-for-service in high-spending areas that might previously have been bound by a 2 percent rate of increase or by a blended national-local rate. How geographic variation in spending is addressed in a universal program is a political issue of the first order.

How Much Redistribution Will The Political System Allow?

Specifying the mandated benefit package and dealing with geographic variations are but two of the many dimensions through which universal coverage involves redistribution. Any universal insurance plan will increase the amount of redistribution from the healthy to the sick, from higher-income to lower-income households, across workers at different firms and possibly within a firm according to their current health benefits, and between owners of different businesses. Ultimately, of course, it is all redistribution across households.

With personal health care services constituting more than one-eighth of the economy, the scope of potential redistribution is large—larger than any non-wartime policy change we can think of, save Social Security. The U.S. political system tends to resist increased redistribution, particularly when it creates identifiable losers, precisely the reason an incremental approach is attractive.¹¹ None of this is to disparage the IOM committee’s call for universal health insurance. We hope that it happens, but we are not holding our breath.

NOTES

1. Institute of Medicine, Coverage Matters: Insurance and Health Care (Washington: National Academies Press, 2001); Health Insurance Is a Family Matter (2002); Care without Coverage: Too Little, Too Late (2003); Hidden Costs, Value Lost: Uninsurance in America (2003); A Shared Destiny: Community Effects of Uninsurance (2004); and Insuring America’s Health: Principles and Recommendations (2004). Joe Newhouse was the National Research Council monitor for most of the volumes in this series.
2. We would impose additional taxes to finance the cost.
3. Given that roughly 15 percent of the population is uninsured at a point in time, the committee’s 3–6 percent value for the increase in total use implies an increase in use among the uninsured of 20–40 percent, a number with which we do not quarrel. Even a larger increase among the uninsured would not change the conclusion that the increase attributable to the uninsured will not be large relative to total spending.
4. The addition of a prescription drug benefit to Medicare can be viewed as beefing up a below-average benefit package.
5. The committee and the literature it surveys make numerous comparisons between averages for the insured and uninsured, which implies a comparison to the average insurance policy.
6. Risk adjustment and potential variation in subsidy rates also arise. See J.P. Newhouse, “Patients at Risk: Health Reform and Risk Adjustment,” Health Affairs 13, no. 1 (1994): 132–146; and J.P. Newhouse et al., “Risk Adjustment and Medicare: Taking a Closer Look,” Health Affairs 16, no. 5 (1997): 26–43.
7. Indeed, the cost-effectiveness calculations in Hidden Costs, Value Lost are suspect because they do not consider either costs or benefits among the insured, including Medicare beneficiaries.
8. A.B. Martin et al., “Trends in State Health Care Expenditures and Funding: 1980–1998,” Health Care Financing Review 22, no. 4 (2001): 111–140.
9. The president’s fiscal year 2005 budget proposal includes a tax credit for the low-income uninsured of 90 percent of premium costs up to $1,000 per adult and $500 per child for up to two children. The credit percentage would phase down for people with adjusted gross incomes between $15,000 and $30,000 and other filing units with incomes between $25,000 and $60,000.
10. This deduction is capped to limit the deduction to interest on mortgages of no more than $1 million.
11. C.L. Schultze, The Public Use of Private Interest (Washington: Brookings Institution, 1977).


Read related papers by Wilhelmine Miller et al., and Henry J. Aaron and Stuart M. Butler.

Click here to view a Webcast of a Kaisernetwork interview with Joe Newhouse.

DOI: 10.1377/hlthaff.W4.179
©2004 Project HOPE–The People-to-People Health Foundation, Inc.