Journal Issue: Health Care Reform Volume 3 Number 2 Summer/Fall 1993
Sorting Out the Costs of Reform
A review of how the costs of health care are distributed under the current system is an appropriate starting point for an analysis of who bears the burden of health care reform. Under the current system, the health care bill is paid by premiums levied on employers and individuals, by private out-of-pocket spending, by government expenditures on Medicare and Medicaid, and by government subsidies to public hospitals and other direct service providers.
In assessing how cost burdens will be shifted by reform, it is critical to distinguish between the costs that society bears as a whole and the new government revenues that will be needed to finance reform. Even proposals that will reduce the share of our economy devoted to health care may require increases in government subsidies (and hence taxes) to ensure that low-income persons have access to the system. Thus, to truly understand the impacts of health care reform, we need to consider both social and government costs.Understanding the Current Burdens
Whether the original source of payment for health care is the federal government (compensating providers for the costs of Medicare beneficiaries), insurance companies or employers acting as insurers, or individuals (as part of their share of employer-subsidized premiums or when no other third-party payers cover the cost of care), the burden of paying for health care falls ultimately on American households. Individuals pay the taxes that support government programs and the costs that businesses incur either through lower wages or lower profits or through higher prices for goods and services. But this cost burden does not fall equally; individuals pay different levels of taxes and pay varying amounts of costs out of pocket.
While it is not possible to trace each individual transaction and to identify exactly who pays, it is possible to analyze the relative burdens across groups of the population. One test of the fairness of any new system of financing health care in the United States will be how it distributes the associated costs across families at different income levels. But first it is useful to focus on the current distribution for families by income.
Figure 1 indicates the results of a recently completed study that attempts to assign the burdens of financing health care by income quintile, that is, any one of the five equal groups into which income receivers can be divided.4 The bottom quintile contains the 20% of families with the lowest incomes in a given year. As shown in that figure, the poorest Americans spent a larger proportion of their resources on health care than any other group. The average burden for all families was 10.1% of income and nonwage compensation; for those in the bottom quintile, it was 13.2%. Families in the highest income quintile used 8.8% of their resources to finance health care. The authors include in these estimates taxes paid that support health care services, the average amount of employer-paid premiums (assuming they are mostly shifted back to employees in the form of lower wages), and out-of-pocket spending on health care services and insurance premiums.
Families with one or more related children are most likely to be found in the bottom quintile. For example, in 1991, more than 60% of all families in the bottom quintile had children, compared with 47% in the top quintile.5 And if we look at young children, 34% of the poorest families included children under age six, while only 19% of the families in the top quintile had young children. Thus, children are more likely to be in families with high health care costs relative to income, which makes them especially vulnerable to the problems associated with the lack of affordable care.
What is not shown in this analysis is that, although members of low-income groups pay a larger share of their income for health care services, those without Medicaid coverage receive less care on average than persons with similar needs but higher incomes.6 Lack of insurance means that many Americans do not get as much care in the early stages of illness, when it can be most efficiently delivered.7
Nonetheless, this distribution of burdens is instructive. We start from a system where households with the lowest incomes spend the greatest proportions of their incomes on health care. Because a major reason for the lack of insurance among those with low incomes is the high cost of coverage, reform proposals that attempt to expand access to care should seek to lower costs.Costs to Society versus Higher Taxes
When looking at the new costs associated with a health care reform proposal, it is important to distinguish between new costs to society (that is, to the overall economy) and new costs to the government. These two concepts are usually quite different but are often confused. Indeed, reform proposals with different impacts on government costs may result in similar costs to the overall economy. Both concepts of financial impact are critical for understanding the debate over health care reform.
Consider the example of a reform that would reduce overall health care spending by generating administrative savings and other new efficiencies. These savings might be enough to offset the expenses of covering persons who are currently uninsured. Indeed, this argument is often made by those who support a Canadian-style single-payer system for the United States.8 But advocating such a change is very different from saying that there will be no need for new taxes and, hence, no new burdens on government. A Canadian-style system requires that people pay enough in taxes to fund most of their health care purchases; in the overall economy, the additional taxes would be offset by lower out-of-pocket spending and elimination of employers' contributions to health insurance premiums for their workers. For the United States, this would mean that a much larger share of health care expenditures would be shifted to the federal government or to federal and state governments. The total government share of health care spending was estimated to be about 45% in 1992.9 If the United States adopted a single-payer system and covered about the same proportion of aggregate spending as Canada, that share would rise to about 75%.10 Thus, even if overall health care spending went down a bit, expenditures by the government sector would increase and result in higher taxes.
Table 1 provides some simple illustrative numbers, starting with estimated system-wide spending of $940 billion.11 While these numbers are approximate, they offer a sense of what might happen if the United States adopted a Canadian-style system that expanded coverage and reduced overall spending. For this illustration, it is assumed that the public sector funds 45% of health care expenditures under the current system and would fund 75% under a Canadian-style reform. Further, it is assumed that administrative and other savings would total $100 billion (up from Woolhandler and Himmelstein's 1987 estimates of $70 to $80 billion8) and that this would be only partly offset by an increase in spending of $60 billion on those who are currently uninsured, an amount extrapolated from 1990 estimates of $45 billion.12 Even though overall health care spending would fall from $940 to $900 billion, the net new public revenues required would total $250 billion.
Some health care reform proposals are popular specifically because they funnel fewer of the costs through the government sector.13 Health care reforms that require employers to provide insurance keep the costs "off the books" as far as government is concerned. (Even here, government expenditures will probably increase to subsidize the unemployed, but are likely to grow only to about 50% of total health care spending.)12 But the costs to society would probably be higher than in the first example. Employer-based reforms would not reduce administrative costs as much as a single-payer approach and, hence, would not be able to fully absorb the costs of expanding coverage in the same way that a single-payer system would.8,13 Savings might result from greater controls over spending and some administrative streamlining once a uniform system is in place. Consequently, this example assumes that society's costs would rise by $20 billion overall. However, the burdens on government would not increase as much as in the first example because business would be asked to pay more. The public sector costs associated with an employer-based health care system are shown in Table 2, and in this instance, government costs rise by "only" $55 billion.
In practice, if one of the goals of health care reform is to expand coverage to the uninsured, many of whom have low incomes and are not employed, it will be necessary to raise some revenues through the government sector even with employer-based reforms. Moreover, in judging the reasonableness of alternative financing options, we should keep in mind both the impact of indirect financing mechanisms, such as higher required employer contributions, and the current burdens of financing health care. These factors determine the distribution of burdens (as shown in Figure 1), and reform of any kind will probably change that distribution. It is simply not possible to avoid some redistribution in the process of providing subsidies to make health care available; to obtain resources to create subsidies, some individuals will be "losers," often in terms of paying higher taxes. Under mandates, some of the losers will be employees if the costs of paying insurance are shifted from the employer to his workers. On the other hand, where employees have had to buy insurance privately on their own, they may be better off even if they must pay the full employer premium share. But many will also be "winners" as well; in fact, this is an important element of efforts to reduce burdens on those with low incomes. These examples apply equally to programs aimed only at children and to reforms for the entire health care system.Estimated Costs of Health Reforms for Children
A very broad range of new government spending on health care reform is possible, depending on what type of reform is implemented and whether it is restricted solely to children. Because the discussion presented above would apply equally well to children-only or children-first programs, we consider both the overall distributional effects of proposals and the new costs through the government sector from these proposals.
For children-only expansions, required new government spending would be relatively modest, especially if reform does not include a full public program. Indeed, this is one reason to begin health care reform with children. For example, an employer-based reform could require that employers either offer coverage to all dependent pregnant women and children or pay a modest payroll tax that would cover most of the costs of the program; the government would cover children in families where the parents are not in the labor force. This approach is usually termed pay or play. In a program of this kind, new government spending could be as low as $6.8 billion (see Table 3). An expansion of Medicaid to cover all children (to age 18) would cost the government more—about $9 billion14-because a pay-or-play approach would require employers to insure children of workers. If only Medicaid coverage is expanded, all low-income children would be in the public program. Costs for a Medicare-type program for all children and pregnant women would be much higher—approximately $55 billion—because all children, including those in higher-income families now covered by a parent's employer-based plan, would be shifted into the public plan. Coverage for all children would flow through the public sector; and as was true in the example in Table 1, government revenues would need to increase substantially to pay for this program.
Estimates suggest that children-only programs would raise government spending by about one-fourth as much as programs for all Americans under age 65 (also shown in Table 3). While more than one-fourth of all persons under 65 years of age are children, children are less expensive to insure; therefore, the federal budget costs of children-only plans are lower. On the other hand, children-only health care programs would require similar administrative structures to those needed for full-population coverage; therefore, the aggregated system-wide costs of administration probably would be larger in a system that treated children and adults separately than in a system which covered everyone. Thus, a children-only program might not achieve all the offsetting cost savings that are often cited for full-population reforms.