The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, its funders.
Mr. Chairman, members of the committee, I appreciate the opportunity to appear before you today to discuss ways to expand health insurance coverage. My name is Alan Weil and I direct the Assessing the New Federalism project at the Urban Institute, a 34-year-old non-profit, non-partisan research institute in Washington, D.C. Before coming to the Urban Institute I was executive director of the Colorado Department of Health Care Policy and Financing, which is the state Medicaid agency.
There is an emerging consensus that public subsidies must be provided to assist the 40 million Americans who lack health insurance. The disagreement that remains centers around the form of those subsidies. Some advocate tax credits, while others advocate expanding existing public programs such as the State Children's Health Insurance Program (SCHIP) or Medicaid.
In my testimony today I will argue that the latter approach—building upon existing public programs—holds far more promise for improving health insurance coverage. The case for tax credits rests entirely on theory and ignores the practical difficulties of providing meaningful coverage in a complex, varied health care system. Existing public programs also have limitations, but they have a 35-year track record of providing comprehensive, cost-effective and stable coverage to those in need.
Tax credits suffer from five problems—problems of availability, adequacy, amount, administration and accountability.
Availability. The most serious problem with tax credits is that of availability. Most tax credit proposals, such as the one offered by President Bush, are designed to encourage people to purchase coverage in the individual health insurance market. Insurers in this market routinely deny coverage to those with any identifiable health problems, or they write coverage that excludes conditions or body systems where there is any history of medical problems. When coverage is offered, rates are many times higher for older adults than for those who are younger. Administrative costs routinely exceed 30 percent. These insurance practices are the only way companies can make money operating in a market where they take on the substantial risk associated with enrolling people with high health care needs.
Most states have adopted regulations—such as guaranteed issuance of policies, guaranteed renewal, and modified community rating— to promote availability in the small group market where employers buy coverage. However, only a small number of states apply these rules to the non-group market.
My purpose is not to advocate a specific set of regulations that should apply to the non-group market, but to highlight the issues involved. There is legitimate controversy about how far in the direction of community rating the non-group market can go and retain its appeal for the younger, healthier population. However, given the current state of the non-group market, regardless of the size of a tax credit, health insurance simply will not be available to those who most need it.
Adequacy. The second problem with tax credits is that of adequacy. The size of the credit—$1000 for an individual and $2000 to $3000 for a family in the President's proposal—does not even cover half the cost of the typical health insurance plan. Analysts agree that few families of modest means can or will pay the balance with their own funds. Tax credit users will primarily end up in plans with cost sharing that runs in the thousands of dollars, with many excluded services, or significant limitations on coverage. Based upon current market practices, we can expect that deductibles will exceed $1000 per person, cost sharing will run 20% or more, and plans will use fee schedules that leave families covering an additional portion of the bill of a large share of health care providers. Covered benefits may exclude, at a minimum, items such as mental health, prescription drugs, preventive care, office visits, and dental care. Coverage limitations may include maximum coverage for hospital care, and limits on how many of some kinds of services a person may use. These limited benefit packages will leave families in exactly the position they find themselves today: deferring needed care because of cost, at risk of bankruptcy if they get sick, and placing a tremendous financial burden of uncompensated care on the entire health care system.
Higher income families may be able to absorb the costs associated with limited health insurance plans. However, families with incomes between $20,000 and $30,000, that are unlikely to have coverage through their jobs and are the target of the tax credit, face denied or delayed care or possible financial ruin if their insurance coverage has these gaps. In addition, all experiments attempted to date show the same thing: most Americans, and particularly those of limited financial means, are simply not interested in bare bones coverage. Thus, even if the tax credit covers a portion of the cost of a health plan, families are unlikely to pay the balance when the coverage it yields is inadequate.
Amount. The third problem with tax credits is that of the amount. Tax credits suffer from the Goldilocks syndrome: no tax credit amount is just right. Everyone agrees a tax credit that is too small will not increase insurance coverage at all. However, a tax credit large enough to help a substantial number of people obtain health insurance is also large enough to draw a substantial number of people out of the employer market, thereby raising premiums for small businesses and shifting costs from the private sector to the taxpayer.
A few analysts have developed complex models to estimate how many people would gain coverage from a tax credit, how many would drop their existing employer coverage, how many employers would drop coverage, and how many currently insured people would benefit financially from the tax credit. Estimates, such as the 6 million who would gain coverage at some point during the year according to analysis by the President's Council of Economic Advisors (CEA), are very sensitive to the assumptions they use. If those models are wrong even by a few percentage points in their predictions about how families, firms, and insurance companies will respond to the new incentives of a tax credit, the number of people who gain coverage after enactment of a tax credit could be much smaller. No matter how carefully we design our models, it is important to understand that in some parts of these models we have very little data on which to base expected responses.
One limitation of all of the dominant models warrants particular attention. Analysts recognize that a fixed dollar tax credit has health insurance purchasing power that varies by a factor of more than five-to-one, depending upon where a person lives. Sophisticated models use actual data on health insurance prices to estimate how many people will buy coverage. However, these models cannot take into consideration the clustered nature of different behavioral responses.
For example, a tax credit of $1000 in a very low cost health insurance market will be enough to permit a young, healthy individual to purchase a fairly comprehensive policy. An individual with annual earnings of about $14,000, or $7 an hour, is probably not offered coverage through his or her employer. Even if that individual is offered coverage, according to the CEA, the employer current tax subsidy would be well under $100. If we believe people will respond to the financial incentives of a tax credit, we must assume that many people in the employee's circumstance would take advantage of the offer of a $900 increase in the tax subsidy, and would move to the non-group market.
By contrast, in a higher cost market, the $1000 credit is not enough to cover more than the most skeletal health plan. In these markets, it is unlikely that more than a handful of individuals will drop their employer coverage to move into the non-group market. The small value of the credit relative to the cost of coverage means there will be little market disruption, but also very few people currently without health insurance who will benefit.
The result is highly clustered, local effects of a tax credit. In one place it will have no effect other than to offer a tax break to those who are already insured. In another, the entire employer market could shift radically, with potential cost increases and loss of coverage for those already insured.
Another often-ignored problem with setting the amount of the credit is how it will interact with existing or potential state policy choices with respect to public coverage through Medicaid and/or SCHIP. The presence of a tax credit large enough to help an individual purchase coverage will also reduce the incentives states have to retain or expand coverage in public programs that require the state to pay a portion of the bill. Faced with the choice between a fully federally-funded tax credit or a matching Medicaid or SCHIP program, states have a clear incentive to rely upon the former. This scaling back of state effort would yield fewer people with comprehensive insurance coverage and a larger fiscal burden for the federal government.
In short, it is impossible to set a credit amount that strikes some theoretically correct balance between helping no one and undermining the existing public and private health insurance system.
Administration. The fourth problem with tax credits is that of administration. At a minimum, a tax credit must be refundable and paid in advance if it is to help a working family purchase coverage. Unfortunately, even with these provisions many families will be unaware of the credit, fail to take advantage of it, or not take it in advance because they will worry they will have to pay the government back if they receive a small wage increase during the year. The existing Earned Income Tax Credit (EITC) provides important evidence. Very few families claim the credit in advance even though it is available. In addition, low-income Hispanic parents—a group disproportionately likely to be uninsured—are less likely to know about the EITC than other low-income parents, and, even those who do know about it are less likely to have received the credit.
Problems of administration arise in part from the desire to use the tax system to effect a goal that is inconsistent with its primary purpose. Although recent provisions, such as the EITC and the child care credit, have included similar features of refundability, neither of those credits involves the same complexity as that of the proposed health insurance tax credit. For example, eligibility for the health insurance credit is based upon the absence of something else—employer sponsored insurance and public insurance—which must be verified. Health insurance is bought by family units that do not necessarily align with tax filing units. In order to have its intended effect of increasing insurance coverage, the health credit must be taken in advance, whereas the EITC can achieve its work-support objective even if filers claim it at year's end.
Accountability. The fifth problem with tax credits is that of accountability. Most people rely upon their employer or a public agency to provide them information about their health plan, assist with problems, and monitor the quality of coverage. But people in the individual market are on their own. If their coverage is cut, their premiums rise, or there is a dispute over their benefits, they must fend for themselves. If the federal government is providing financial incentives to purchase coverage, individuals will expect the government to make plans available for review. Consumer outcry among those who are denied coverage or who feel mistreated by their health plan will create immense pressure for the federal government to do something.
Solving these problems. Some of the problems with tax credits can be solved, but only at the cost of exacerbating others. For example, the availability of coverage could be partially solved if the federal government imposed insurance market reforms on the states. However, this solution would restructure the federal/state relationship in regulating the health insurance market in a manner that would threaten accountability, and would prompt a contentious battle over federalism. Similarly, coverage could be made more available if, for example, the tax credit could be applied toward the purchase of coverage through an employer. However, this solution would dramatically increase the cost of the proposal, meaning the amount of the credit would have to be reduced so much that it would not provide any meaningful coverage. The problem of amount could be solved in part if the size of the tax credit were more closely calibrated to match the cost an individual or family would face purchasing coverage in the market. However, this solution would require a much more complex system of administration, gathering information on peoples' age and health status that is not consistent with the current tax code.
Unfortunately, different proponents of tax credits emphasize different aspects of their proposals, seeming to suggest that all of these problems can be solved. Yet, if one looks closer, it becomes clear that the different features of these proposals are often in conflict with each other, and would yield substantial increases in the problems faced in other areas.
A note on the non-group insurance market. Much of the conflict over tax credits is a conflict over different perceptions of how well the non-group insurance market functions. It often seems that people are talking past each other. A bit of clarification of terms may help.
In health insurance, as in other kinds of insurance, there are two kinds of risk to consider. One is random risk—the fact that some percentage of people will have an accident or an unanticipated event during any given period. The other is systemic risk—the fact that some people have health conditions that make the expected cost of their needed health services higher than others. Defenders of the current non-group market claim that the market is effective in spreading risk, but they are focusing entirely on the first kind of risk. It is true that, if a group of fairly young, healthy people all purchase coverage in the non-group market, the random health risks they face will be shared across the group. The problem with the non-group market is that it does not spread systemic risk. Therefore, the older pay more than the young, and the sicker pay more than the healthy. This stands in contrast to the practices of employers, where most people obtain their coverage, where systemic risk as well as random risk is spread across covered employees.
It is certainly true that there are problems with spreading systemic risk in a voluntary non-group market. If the risk is spread too far, the young and healthy will not purchase coverage because of its cost. However, pushing a large number of people into the non-group market, where they bear the full cost of their systemic risk, will change the conception that most Americans currently have of insurance coverage.
Public programs. Existing public programs have their limitations, but they also have a 35-year track record. They provide real, comprehensive, cost-effective stable coverage. They target spending on those most in need, and they minimize incentives for the private sector to drop coverage. Public programs have gone through a positive transformation in recent years. They have simplified applications and enrollment processes, crafted new market-based benefit packages, improved education about coverage options, and have been tackling old problems like how to assure access to critical services like dental and mental health care. Public programs face challenges and they are not and should not be for everyone. However, their strengths and weaknesses are known, and they can be modified to meet the needs of specific populations. They contain systems of accountability and they are up and running.
Projections of how many people would gain coverage if public programs were to expand further are not certain. But those projections can be made with far more confidence than is the case with tax credits.
Conclusion. At a time when the nation faces tight fiscal constraints and growing numbers of uninsured, it is essential that limited resources be spent where they will be most effective. While some models project many people will gain insurance coverage if a new tax credit is adopted, these models are based on many layers of assumptions, few of which have been tested. If the goal is to reduce the number of people without health insurance, spending money on tax credits is a huge gamble paid for with taxpayer funds. It is using positive results from a computer model to justify a new, multi-billion dollar entitlement. By contrast, states were poised to make substantial progress on the issue of health insurance until fiscal circumstances recently took a sharp turn for the worse. Even a modest expenditure of federal funds could revive the state and local creativity we observed just a year ago. This would be an expenditure based upon a track record, not on a theory and a model.
Latest Research
Browse by Topic:
Printer-friendly summary
