June 16, 1999

June 16, 1999 | Testimony on Health Care

A Heath Tax Credit to Assist the Uninsured

Testimony before the House Committee on Ways and Means - United States House of Representatives

Mr. Chairman, my name is Stuart Butler. I am Vice President for Domestic and Economic Policy Studies at The Heritage Foundation. The views I express in this testimony are my own, and should not be construed as representing any official position of The Heritage Foundation. Some of the following material is taken from a forthcoming article I have co-written with David Kendall of the Progressive Policy Institute. Nevertheless, some of the conclusions I draw differ from our joint position, and thus do not necessarily represent his views.

I am pleased that the Committee on Ways and Means is giving consideration to incorporating some form of health tax credit into the tax code. There is growing support outside Congress for introducing changes in the tax code to make it more rational concerning health expenditures and to help the uninsured and to help the uninsured, and these proposals often include a tax credit component. Organizations favoring tax-based reforms include the American Medical Association and the National Association of Health Underwriters as well as scholars in such research organizations such as The Heritage Foundation, the Urban Institute, the American Enterprise Institute, the Cato Institute, and the National Center for Policy Analysis. Moreover, many Members of this Committee, as well as those of other House and Senate Committees, have either introduced health tax credit bills or are considering such legislation. With this growing interest in the approach, I believe the time is ripe for Congress to act this year on a health tax credit.

The interest in introducing a health care tax credit stems from two related features of the current health care system. First, there is growing recognition that the current employer-based system (which is heavily subsidized by the tax exclusion for employer-sponsored health insurance) is a highly inadequate vehicle for providing health coverage in certain sectors of the economy. A new health insurance tax credit would help to stimulate the creation of a parallel health insurance system for working people who are not well served by employer-sponsored insurance. Second, the support for a tax credit (rather than, say, a widening of the exclusion or the introduction of a deduction) recognizes the inefficiency and ineffectiveness of the tax exclusion as a device to help Americans to afford health care.

The growing level of uninsurance in this country underscores the need for at least modest steps, and its causes reinforce the belief that a tax credit would be a sensible step to take. As the Committee is well aware, the number of uninsured individuals is over 43 million, with uninsurance reaching epidemic proportions in some communities. Approximately one-third of Hispanic-Americans are uninsured, for example, as are about one-half of the working poor. Significantly, the uninsured are predominantly within working families; only about 16 percent of the uninsured are outside such families. And even though 24 percent are in families with workers employed part-time for part of the year, 60 percent are in families with an adult working full-time yearround.

Surveys indicate that about 75 percent of the uninsured say they simply cannot afford coverage, or they have lost coverage that was once available through their employer.

Although millions of Americans enjoy the certainty of good, predictable coverage through their place of work, it is becoming increasingly clear that the place of employment is not an ideal method of obtaining coverage for many Americans, particularly in the small business sector. Unfortunately, current tax policy is heavily biased against any other method of obtaining coverage.

Consider the following:

  • In an economy with increased job mobility, for an ever-larger proportion of the population an employment-based group is no longer a stable, long-term foundation for health insurance. Even if a family can expect to receive coverage whenever the main earner changes jobs, typically there will be some change in the benefits available or the physicians included in the plan. The higher the degree of job mobility for a family or in an industry, the higher the degree of change and uncertainty associated with employment-based health insurance.

  • Although major employers, with a large insurance pool and a sophisticated human resource department, may be considered a logical institution through which to obtain health insurance, this is not the case with most smaller employers. These employers typically lack the economies of scale, and usually the expertise, to negotiate good coverage for their employees, and it should be no surprise that uninsurance is heavily concentrated in the small business sector. In 1996, just under half the firms with fewer than 50 employees offered insurance, while the figure was 91 percent for those with 50 to 99 employees and 99 percent for those with more than 200. For those firms with fewer than 50 employees in which most workers earned less than $10,000, only 19 percent were offered health benefits. Furthermore, Hay/Huggins has found that, in 1988, average administrative costs exceeded 35 percent of premiums for firms with fewer than 10 employees, compared with 12 percent for firms with over 500 workers.

  • Other large, stable groupings exist that could sponsor health insurance, but these are discriminated against in the current tax system. For example, unions could carry out exactly the same functions as an employer regarding health insurance. Indeed, the Mailhandlers Union and other unions or employee associations act as plan organizers in the Federal Employees Health Benefits Program (FEHBP). But union-sponsored plans are quite unusual outside the federal government because enrollees in union-sponsored plans typically are not eligible for the tax benefits associated with employer-sponsored insurance. Yet many workers who have only a loose affiliation with their employer, or work for smaller employers who do not provide insurance, have a long-term, close connection with their union. Moreover, the union would be a very large potential insurance pool. Similarly, large religious organizations, such as consortia of churches in the African-American community, would be a far more logical vehicle from which to obtain health insurance, thanks to the size of the insurance pool and the sophistication of the church leadership, than most of the businesses employing members of such churches. Yet again, the tax system is biased against these alternatives.

How the Tax System Exacerbates Failings of Employment-based Coverage

Under the current arrangement for working-age families, employees receive a tax exclusion if they allow their employer to allocate part of their compensation for a health insurance policy owned by that employer. This arrangement helps cause uninsurance in several ways. For example:

  • Because this tax exclusion is available only for employer-sponsored coverage, a working family without employer-sponsored insurance has no subsidy through the tax code to help to offset the cost of buying its own coverage or health care. Thus families who lose their work-based insurance for any reason, such as cutbacks in benefits or jobs by the employer, suffer a double blow -- not only do they lose the insurance, but they also no longer receive a tax subsidy to pay for care. Not surprisingly, high degrees of uninsurance are prevalent among working families with moderate and low incomes

  • The tax benefits available for employer-provided coverage are a very inefficient method of helping low-income workers to afford care. Because compensation in the form of employer-sponsored insurance is excluded from an employee's taxable income (avoiding payroll taxes as well as federal and state income tax), by far the largest tax benefits go to more affluent workers in the highest tax brackets. Those at the lowest income levels (especially those who do not earn enough to pay income taxes) receive little or no tax subsidy. According to John Sheils and Paul Hogan (Health Affairs, March/April 1999), the value of the tax exclusion in 1998 was over $100 billion at the federal level (including income and payroll taxes) and an additional $13.6 billion in relief from state and local taxes. While the average tax benefit per family was just over $1000, the tax benefits were heavily skewed toward higher-income families. Sheils and Hogan estimate that families with incomes in excess of $100,000 benefited to the tune of an average of $2,357, while families with incomes of less than $15,000 received benefits worth an average of just $71 (although this includes uninsured families' receiving no tax breaks at all). Some 68.7 percent of all the tax benefits in 1998 went to families with incomes in excess of $50,000.

How a Tax Credit Would Help

Introducing a tax credit for health expenditures for families that lack employer-sponsored insurance would begin to rectify the deficiencies in the current tax system and, in doing so, would begin to stimulate the provision of health insurance through organizations other than employers. Non-employer sponsored coverage would not be intended to replace successful company-based plans, but could provide an alternative for families that do not have access to insurance through their place of work, or that have clearly inadequate or inappropriate employer-sponsored coverage.

A tax credit would have three key benefits. First, it would be worth at least as much to lower-income families as upper-income families -- unlike the tax exclusion, which is worth far more to people in higher tax brackets. Second, it could be made refundable at least against payroll taxes in addition to income taxes. This means workers without an income tax liability could still claim the credit, thereby providing some help to nearly all the uninsured. In contrast, an individual tax deduction for health insurance has the potential for reaching only about one-third of the uninsured because it would not be refundable and many low-income workers do not have any income tax liability. Third, a credit would be available regardless of job status and would make coverage more affordable for workers between jobs.

Various credit designs proposed in recent years possess these key features. Credit can be a fixed amount or can vary according to a variety of factors, including a worker's expenditure on insurance, income, demographic and geographic factors, and health risks. Major tax credit proposals in the past have ranged from a sliding-scale credit based on income and health expenditures, such as the bill introduced in 1994 by Senator Don Nickles (R-OK) and Representative Cliff Stearns (R-FL), a fixed-sum tax credit such as the bill introduced recently by Representative John Shadegg (R-AZ), or a percentage credit against costs, such as the bill introduced by Representatives James McDermott (D-WA) and James Rogan (R-CA).

The McDermott-Rogan proposal would provide a refundable tax credit for 30 percent of a family's expenditure on health insurance, which is based on the value on the current tax subsidy for a taxpayer in the 15 percent income tax bracket plus the exclusion from Social Security's Federal Insurance Contribution Act (FICA) taxes. The Shadegg proposal would provide a dollar-for-dollar, refundable credit of up to $500 for individuals ($1,000 for families) for the purchase of health insurance.

These different forms of tax credit have subtly different effects. For example, a tax credit for a given percentage of the cost of insurance could encourage overspending by some families, just as the current open-ended subsidy does in employment-based coverage, although this latter effect is reduced if there is an income cap or if the total credit is capped. A simple percentage credit also could leave low-income people still unable to afford coverage. On the other hand, a tax credit for a fixed sum of health care coverage could concentrate the most help on the needy and encourage spending only up to that amount. That would minimize overinsurance, but families facing high costs would incur the full marginal price of needed extra services or coverage mentioned earlier. For workers with a serious health condition facing higher premiums, the ideal tax credit would be a sliding scale credit adjusted upward according to the ratio of cash and income. Such a credit would have the need to subsidize higher-risk workers through community rating laws that perversely benefit high-income, high-risk workers at the expense of low-income, low-risk workers. Most states permit insurance premiums to vary at least somewhat according to health risks and demographic factors in both the individual and small group markets, the two markets mostly likely to be affected by a refundable tax credit. Thus, a tax credit for a percentage of spending (especially a sliding scale credit) would take better account of these differences.

A fixed or percentage tax credit could be provided without regard to income. But clearly that would mean a lower degree of assistance for the poor -- for the same total revenue cost -- than a targeted credit. A tax credit targeted toward those who can least afford coverage, however, would mean there must be some form of phase-out based on income. Such phase-outs necessarily create higher effective marginal tax rates for taxpayers who fall in the phase-out range. This problem is especially pronounced for certain low-income workers, who can face marginal tax rates of 100 percent or more due to the phase-out of several income-based programs such as the earned income tax credit, welfare, day care and Medicaid subsidies, housing subsidies, and food stamps. This problem occurs with any subsidy arrangement, of course, not only with tax credits. More sweeping tax credit reforms, such as the Nickles-Stearns bill, resolved this to a large degree by changing the entire tax treatment of health care, thereby permitting a very gradual phase-out of the credit.

Some Questions and Answers on Health Tax Credits

Q. Would a tax credit undermine successful employment-based coverage?

A. Not if designed properly. It would help to provide an alternative with the characteristics of successful employer-sponsored plans for those currently outside the employment-based system -- such as large, stable group insurance pools and administrative economies of scale. But it is important that prudent steps be taken to combine a credit with a "wall-of-separation" strategy to limit the probability that successful employer plans would be dismantled, either because of a decision made by the employer, or because individual workers preferring the credit undermined the firm's insurance pool.

Certain design elements could be incorporated into a credit to minimize the risk to good employer-sponsored plans. For example, the credit might be made available only in such cases that insurance is not available from the employer.

Q. Would workers with little cash be able to front the cost of insurance before they could claim the tax credit?

A. Yes. The idea that a tax credit means employees would have to wait until the end of the year to obtain tax relief is a myth. Just as mortgage interest tax relief, or a child care credit, is obtained by most families over the whole year by an adjustment to their withholdings, the same would be true of a credit. In addition, a novel idea proposed by Senator Tom Daschle (D-SD) would simplify things even further for many families. Senator Daschle would let the insurer reduce its own tax withholdings for each person who voluntarily assigns the value of his credit to that insurer for the purpose of purchasing health insurance. This approach could be particularly helpful to the unemployed if it applied to Consolidated Omnibus Budget Reconciliation Act (COBRA) plans as well.

Q. Would a credit be an efficient way to provide help? Wouldn't much of the money devoted to financing the credit actually go to people who are already buying insurance?

A. While millions of families are uninsured, there are many families who lack employer-based coverage but have decided to purchase their own insurance, typically with no tax relief. Clearly, these families would immediately take advantage of any available tax credit to offset the cost of their current coverage and/or improve it. For this reason, a significant part of the revenue cost of a tax credit would go to these families, meaning that only a portion of the revenue costs would be used by uninsured families to obtain insurance. Depending on the size of the credit (the larger the amount, the more likely uninsured families would be to take advantage of it), the proportion of "tax expenditures" leading to actual reductions in the uninsurance rate could vary widely.

Some critics of the tax credit approach conclude that a tax credit would be "inefficient" in that many people who today buy their own insurance would simply use the credit to offset their cost without increasing their coverage. But this presupposes that equity is not an appropriate objective, in part, of the tax credit strategy. Yet one of the aims should be to make sure that people of similar circumstances receive the same help, and that it should not be considered a policy flaw if tax relief is provided to families who have saved elsewhere in their household budget to pay for coverage today.

Q. Would a credit be large enough for low-income people to afford coverage?

A. To be sure, studies and surveys suggest that millions of low-income Americans still would consider coverage to be prohibitively expensive even with a refundable tax credit of, say, 30 percent. This observation is used to argue that the credit approach would be ineffective. But a tax credit approach should not be seen in isolation as a complete solution for all the uninsured. Other subsidies and programs exist and are needed -- but these other approaches are more likely to be successful if a family can add part of the cost through a federal tax credit. In particular, states have been using their own and federal resources (such as Medicaid and the State Children's Health Insurance Program, or SCHIP) to provide assistance to families needing health insurance. A refundable federal tax credit of, say, $1,000 for a family should be seen as a foundation on which to build with these other programs and resources. A $1,000 credit would mean that we are $1,000 closer to financing the cost of insurance for a family.

Q. Can other organizations really be as effective as employers in organizing coverage?

A. Yes and no. Many large corporations today have the sophistication, scale of buying power, and presence in the community to outperform any other group in organizing good, economical health coverage. In the system envisioned by the authors, these would be the logical vehicles for coverage, at least if employment tended to be long-term in the firm. Moreover, a tax credit system could also allow families to buy into the health plans of corporations for whom they do not even work, if this made sense for the corporation. Many large firms have made the decision to turn an internal service into a profit center for outside customers. The Sprint telephone company, for example, grew out of the internal communications system of the Southern Pacific Railroad. And John Deere & Co. spun off its health benefits operation as a health maintenance organization in Iowa. If families could obtain tax relief to buy coverage outside their own firm, one would be able to imagine large corporations with huge health plans deciding in the future to offer a competitive insurance service to non-employees.

In many situations, non-employment based groups would have a comparative advantage and would be more logical and skilled organizations. Moreover, these groups are not merely potential pools for coverage. In many instances, they have a "community-of-interest" connection with families, which means they could be expected to work for the long-term interest of these families. Consider, for instance, the potential of union-sponsored insurance in the restaurant and small hotel sector. In this sector, firms tend to be small and employee turnover high, while unions are available that are large and sophisticated. Unions in general have considerable expertise in bargaining for health care and would be the health care sponsor of choice for many Americans -- even for those who do not wish to be active union members. Within the FEHBP, for instance, the Mailhandlers Union provides coverage to many federal workers who join the union as associate members merely to avail themselves of the health plan.

Groups of churches in the African-American community also could be preferred sponsors of care in a system in which subsidies and tax benefits were not confined to employment-based plans. In many communities served by these churches, employers are small and employee turnover is high, yet families have a strong and continuous affiliation with the church. Moreover, America's black churches have a long history of serving the secular as well as the spiritual needs of their congregations by providing housing, education, insurance and other services.

To be sure, there are legitimate concerns to be addressed in considering the role of such organizations in health care. One is the stability of the insurance pool -- if individuals can easily affiliate or end their affiliation it may be difficult to secure coverage without wide price fluctuations over time (of course, this is also a problem with small employer pools in some industries). Another, linked to this, is the worry that adverse selection may undermine the group.

It is unclear how large these problems are. In the FEHBP, for instance, many plans operated by organizations (such as the Mailhandlers mentioned earlier) allow individuals from outside the base group to affiliate for a small fee simply to obtain coverage, and all enrollees are charged the same community rate. Yet the groups are surprisingly stable, perhaps due in part to the relatively high costs for individuals to calculate and make plan choices based on their own predictions of their own health care costs. Yet even if stability and adverse selection is accepted as a serious concern, steps at the state or federal level could be taken to increase the stability of the group. For instance, there could be waiting period after joining the group before the family could join its health plan. In addition, one-year minimum enrollment contracts could be required. Another protection might be to place a minimum requirement on the membership of the pool, which might be achieved through a multi-year consortium of several churches, say, to make the pool large enough to withstand the inflow and outflow of members. The groups also could operate under insurance pooling and rating requirements developed by states.

Q. Would a health care tax credit be a further impediment to tax reform?

A. In a simpler, flatter tax system, there would be no tax preference at all for health expenses. If the current tax expenditures for health care were to be used to help "finance" an across-the-board rate reduction, it could significantly lower the rates in a flat income tax or sales tax, which would of itself make health insurance more affordable.

If, however, it is assumed there is little prospect of eliminating the tax preference for health costs, a tax credit -- especially a credit of a fixed amount per family -- would be reasonably consistent with tax simplification. If over time, the tax treatment of health care were gradually shifted from today's exclusion and deduction system to a credit, this would be more compatible with a flat tax or sales tax than the current system. The reason for this example, the health tax credit could be subsumed into the general exemption for families in a flat income tax.

Growing rates of health uninsurance in the United States are unacceptable and will lead to steadily rising pressure on Congress to take action. After recognizing the root causes of this problem, which lie in the combination of a tax bias toward employer-sponsored insurance and the inadequacy of that insurance system in certain sectors of the economy, it would be prudent for Congress to move quickly but carefully to correct the problem. A limited tax credit for expenditures on insurance not provided through the place of employment is a sensible step that Congress could take this year. It would not mean a radical drop in the number of uninsured, unless there was a very large commitment of funds, but it would be an important first step helping the uninsured and to achieving the general reform of tax benefits for health care. It would also stimulate the creation of parallel institutions that would sponsor insurance in those sectors of the economy in which employers are a very inadequate vehicle for coverage. But if Congress does not take the first step this year, when federal finances are in surplus and the economy is strong, it is likely to face far more difficulties in taking a step in the future if the economy weakens and deficits return.

Stuart Butler, Ph.D. is Vice President of Domestic and Economic Policy Studies at The Heritage Foundation.

About the Author