February 15, 2001


by Iris J. Lav and Joel Friedman


View PDF version

If you cannot access the file through the link, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.

Although the majority of Americans under age 65 have health insurance, one in every six of the non-elderly — some 42 million people — are not insured. Recently, President Bush, various members of Congress, and some policy organizations have proposed addressing this problem through the creation of refundable tax credits with which individuals could purchase insurance.

Helping uninsured people obtain health coverage is an important goal and an appropriate use of public funds. Public funds devoted to this purpose should be used efficiently and effectively. Unfortunately, the research and evidence suggests that the type of individual tax credits now under discussion are unlikely to help many uninsured families obtain quality health care and could erode coverage for substantial numbers of currently insured families. Among the problems and shortcomings of the individual tax credit proposals now circulating are the following:

The type of individual tax credits for purchase of health insurance that are currently being discussed represent a costly policy that is unlikely to result in adequate health coverage for most individuals and families that currently are uninsured. Other policies, such as expansions of public programs — perhaps coupled with other selected tax-based strategies such as employer tax credits — hold substantially greater promise for covering the uninsured.


Value of the Credit is Typically Too Low to Make Health Insurance Affordable

To be successful in substantially reducing the ranks of the uninsured, a program to cover the uninsured must make insurance affordable for the low- and moderate-income families that make up most of the uninsured population in the country. Nearly 60 percent of uninsured people live in families with incomes below $30,000, and about 70 percent live in families with incomes below $40,000. The individual tax credits that have been proposed are too small to make health insurance affordable for most of these families; research shows that perhaps only one-fifth of moderate-income uninsured families, and a still-smaller percentage of low-income families, would be able purchase insurance with credits of the size under consideration.


Tax Credits Would Stimulate Low-Quality Health Insurance Policies

Some proponents of tax credits argue that some health insurance is better than no health insurance. Even if the amounts of the proposed credits would be inadequate to allow purchase of a more comprehensive policy, the argument goes, families would be able to use tax credits to purchase some type of insurance policy. Indeed, it is reasonable to expect that companies would develop and market policies that people could buy using the credit alone, without adding additional funds. Enactment of a credit for purchase of health insurance of $2,000 per family, for example, would likely result in family policies being offered for sale that cost about $2,000. Such insurance policies clearly would have limited benefits; the question is whether those benefits would be of significant value to families. The evidence suggests that these limited policies often would not provide families with meaningful coverage to address their health care needs.

2-15-01tax2-f1.jpg (17889 bytes)


The Individual Market is Risky, and Credits Could Degrade Coverage for Older, Less Healthy Workers

Most of the proposed tax credits would subsidize health insurance purchased in the individual, non-group market, as opposed to the group market available to employers. Non-group insurance is more expensive than comparable group insurance, reflecting the high administrative costs associated with marketing and selling policies to individuals and the inability of non-group insurers to pool risk over a large group. Moreover, policy premiums in the individual market can vary widely, depending on a person's age, gender, health status, and geographic location. In some areas, non-group insurance is simply not available to people who need it, as insurers can deny coverage to applicants who have preexisting health conditions. In most states, the non-group market is weakly regulated.


Individual Tax Credits are Not Well Targeted on the Uninsured

Individual tax credits are an inefficient way to cover the uninsured. Tax credits for individuals tend to provide more subsidies to people who already have insurance than to the uninsured. Research indicates that most of the tax credit proposals would provide the bulk of their subsidies to individuals and families who already are purchasing insurance on the individual market, or to employees whose employers drop coverage as a result of the availability of the individual tax credit.

Professor Jonathan Gruber of the Massachusetts Institute of Technology has conducted research that shows the extent to which individual tax credits would result in churning in health insurance coverage — that is, workers moving from employer coverage to individual coverage with a credit — rather than result in large numbers of previously uninsured individuals and families becoming insured. Gruber found that an individual tax credit that is similar to the credit in most of the major proposals being considered would have only a modest impact in reducing the number of uninsured, because three-quarters of those making use of the credit would already have health insurance — either through their job or purchased on their own.

About 15 percent of all uninsured people — or six million people live in families where the worker has declined employer-sponsored coverage. Surveys have shown that the most common reason for the uninsured to decline employer-based coverage available to them is that it is too expensive.

A tax credit that is available to everyone regardless of whether they are eligible for employer-offered insurance also has the potential to erode the existing employer-based system. For example, healthy employees who could get a better deal on the individual market would not have the option of leaving the employer plan if the credit were restricted to those whose employer does not offer insurance. Credits that are available regardless of the availability of employer coverage, however, could set off a spiral of "adverse selection" as young and healthy workers use the credit to buy non-group insurance at a lower cost than they pay for employer-sponsored coverage.

The Progressive Policy Institute Proposal

A proposal by the Progressive Policy Institute attempts to grapple with some of the more problematic aspects of an individual tax credit. In particular, it attempts to develop a way to provide individual tax credits while avoiding some of the vagaries of the non-group insurance market. While a thoughtful effort, the PPI proposal still exhibits many of the pitfalls and problems of individual tax credits identified in this analysis. Moreover, the solution that it develops to mitigate non-group insurance problems relies on requiring all employers to offer and process (although not subsidize) insurance for their employees. It also mandates states to create insurance pools and products. Absent these requirements and mandates, which are not likely to be acceptable to smaller businesses and states and thus are unlikely to be approved by Congress, the PPI plan is quite similar to the other proposals discussed in this report.

For taxpayers not covered by an employer-subsidized plan, the PPI tax credit would be $1,000 for individuals and $2,500 for families. A smaller credit would be available to employees using an employer-subsidized plan. (Note that PPI's explanation of its plan is unclear with respect to when employees would be eligible for the larger or the smaller credit. The proposal appears to say that a taxpayer who is eligible for an employer plan but prefers to purchase individually would get the larger credit.) Like the credits in most of the other plans currently under discussion, the PPI credit would not be large enough to make insurance affordable for the low- and moderate-income families that comprise most of the uninsured.

Even if the PPI plan were successful in its intention to encourage group purchase and thereby hold down the cost of insurance as compared to the non-group market, the credit would make insurance affordable only for a modest fraction of the uninsured. The cost of family coverage averages $6,350 per year in the group market, of which $2,500 could be paid by the credit. That would leave a family earning $30,000 with the need to expend one-eighth of its gross income to purchase insurance. For most families at this income level, costs would remain prohibitive.

The PPI plan tries to create access to group insurance for those who currently do not have such access. The proposal requires employers to offer insurance plans to their employees, either through employer-arranged group insurance or through special groups set up by states. The proposal also requires employers to determine which of their employees are eligible for the credit, based on family income. Furthermore, the proposal asks employers to advance to their eligible employees over the course of the year, in their paychecks, the amount of their anticipated credit. The advance payments are intended to alleviate the problem of moderate-income families being unable to afford to pay up-front for insurance and then wait for a refund a year later at tax-filing time. For the PPI plan actually to garner these advantages and be an improvement over tax credit plans that rely on the individual insurance market, however, legislation would have to be enacted requiring employers to perform all the mentioned functions: offer insurance plans, enroll their eligible employees, and provide advance payments of the credit through employees' paychecks. Nearly half of uninsured workers are employed by small businesses, which do not have a history indicating eagerness to take on these roles.

Even if the employers were required to perform these roles, there may be barriers to their doing so. Determining which employees are eligible for a credit that is based on family income and providing advance payments are far more complicated matters than they employees' wages, not their employees' family income. Employers could ask their employees about family income at the beginning of the year, but the total family income of low- and moderate-income families often fluctuates considerably during a year due to changes in family situation, job losses or job changes, overtime pay, and a number of other variables. For example, a Census study shows that only one-third of people who were poor in a given month were poor for two full years.

If an employee receives a credit in his paycheck during the year but finds at the end of the year that his income exceeds the maximum eligibility level for the credit, he would owe large payments to the government at tax time to repay the credit. There are major questions about how the IRS would recapture these overpayments, since many low- and moderate-income employees would not have the funds to repay the credit. Would employers be required to withhold and remit to the IRS additional amounts from employee paychecks in the subsequent year to repay the credit? Would the risk of owing a repayment deter employees from using the credit? Problems with predicting annual family income are a major reason that only one percent of EITC recipients take their credits as advanced payments, even though — in an attempt to avoid these problems — EITC advance payments cover no more that 60 percent of the EITC benefit for which a family is expected to qualify. These problems would be magnified with respect to a health insurance credit since, to be effective, a health credit would require advance payments to be widely used and — given the large gap between the credit and the cost of insurance — would probably require advancing 100 percent of the credit during the year.

In addition, the PPI proposal imposes new mandates on states. Under the PPI plan, states would be required to create risk pools and purchasing groups that would help small businesses offer insurance. States have had limited success in this area, even though several have tried. A recent review of state purchasing alliances by the RAND Corporation, for example, concluded that these purchasing arrangements did not reduce the cost of insurance to the small businesses that used them or increase the percentage of employers that offered insurance although they did help the small employers offer a wider range of policies. States may be reluctant to accept new mandates in an area in which their previous efforts have been discouraging.

The PPI plan relies on the premise that mandates on states to create purchasing groups will hold down costs and make insurance more affordable. Another aspect of the PPI plan, however, could drive up the cost of insurance. The PPI plan appears to allow employees to choose among health plans, including plans other than those offered by their employer, and still receive the full tax credit. (The details of how this would work are not clear in the PPI materials.) Thus, an individual employee might choose to take a plan offered by his employer, the state, or perhaps a plan offered on the non-group market, using the tax credit to help pay the premiums. This creates the potential for "adverse selection" by tempting younger and healthier employees to look for lower-cost coverage outside of the employer-offered plan, leaving older and less healthy employees in the employer plan. To the extent that older and less healthy employees become concentrated in the employer-offered plan while individuals with less need for health care choose other forms of insurance, the cost of premiums for insurance through the employer plan would rise and become less affordable for those most in need of insurance.

Finally, the PPI plan is particularly problematic for families with children covered under Medicaid and SCHIP. Instead of expanding Medicaid and SCHIP to cover low-income working parents, the PPI plan would either cover such parents alone or lead parents to move their children from comprehensive SCHIP coverage to what would in most cases be less comprehensive and more expensive insurance purchased with the credit. Under either choice, the family would be less well off than if SCHIP were expanded so more states would use SCHIP funds to cover parents along with their children. Some 17 states already cover parents with incomes up to at least 100 percent of the poverty line under Medicaid and SCHIP, and several others are moving in that direction. To the extent that the PPI plan discourages future state expansions of this nature, it could be a step backwards.

If a tax credit encouraged a significant number of younger, healthier workers to switch out of employer-based coverage, this would leave primarily the older, less healthy workers in the employer's insurance pool, resulting in higher average premiums for those who remain. As adverse selection drove up the costs of employer-provided insurance, over time less and less insurance would be provided through the group market by employers. This could raise the cost of health insurance generally, and expose even more people to the problems in the individual, non-group insurance market.


Credits Will Be Complex to Administer and Implement

There are many administrative complexities associated with individual tax credits, some of which would likely undermine the effectiveness of a credit.


Building on Public Programs and Employer-Sponsored Insurance is a More Effective Way to Help the Uninsured

With the introduction of the State Children's Health Insurance Program in 1997, states have extended insurance coverage to a wider range of children and parents, moving beyond the exclusively poor population traditionally eligible for Medicaid. About 3 million children are now covered through programs supported with SCHIP funds. The total number of uninsured children in the United States fell by more than one million between 1998 and 1999, in part because of increased enrollment of low-income children in publicly funded programs. Census data show a large effect on low-income children above the poverty line.

While coverage for parents of eligible children is more limited, a growing number of states are expanding coverage for working parents. Research shows that when eligibility of low-income working parents is expanded, and children and parents can be covered together through a common policy, enrollment of eligible children increases substantially. Currently, 94 percent of uninsured low-income children are eligible for SCHIP. This strongly suggests that increasing funding for SCHIP to expand coverage to parents of eligible children would bring more of these eligible, but not yet enrolled, children into the program. As HHS Secretary Tommy Thompson said in his confirmation hearing, "You know, the biggest problem, CHIP does not allow the parents to get health insurance at the same time. And so without allowing parents to enroll into the SCHIP program, you are going to, I think prevent a lot of children from being enrolled and a lot of working poor not being able to." Expanding eligibility to parents of covered children, and possibly to other low-income uninsured adults, would be an effective and efficient next step.

As noted, individual tax credits rely on the individual, non-group market that has many problems that cannot be overcome without a level of regulation unlikely to be acceptable to many proponents of tax credits. Building on and expanding Medicaid and SCHIP avoids those problems.

For uninsured populations that would not be reached by a Medicaid/SCHIP expansion, tax credits to encourage employers to offer health insurance, perhaps limited to businesses with lower-wage workers or to small businesses, offer an approach that is less subject to the problems of an individual tax credit. An employer credit would build on the existing system of employer-sponsored insurance, with its advantages of risk pooling and lower prices. It would avoid a fundamental problem with most of the individual credit proposals, which require people to purchase insurance in the precarious non-group market where the older and less healthy, in particular, face higher prices and restricted availability of coverage.

An employer credit also could facilitate greater take-up by eligible low- and moderate-income families. It does not have the timing problems that individual credits have in getting the subsidy into the hands of cash-strapped families when they have to pay their insurance premiums. Advance payment schemes, which are unlikely to be effective based on experience with the EITC, are no longer needed when the employer is purchasing the coverage.

Employer credits also have some limitations. For example, employer credits — like individual credits — are relatively inefficient because they would give credits to employers that already provide insurance. Nevertheless, on balance employer credits are better policy than the individual credits under consideration.

Overall, expansions in public programs, perhaps coupled with an employer credit, offer a much greater chance of reducing the number of Americans without health insurance than do the individual credits being considered.



Broadus, Matthew and Leighton Ku, "Nearly 95 Percent of Low-Income Uninsured Children Now Are Eligible for Medicaid or SCHIP," Center on Budget and Policy Priorities, December 6, 2000.

Center for Studying Health System Change, "Who Declines Employer-Sponsored Health Insurance and Is Uninsured?," CSHSC Issue Brief Number 22, October 1999.

Council on Economic Advisors, "Reaching the Uninsured: Alternative Approaches to Expanding Health Insurance Access," September 2000.

Duchon, Lisa, et. al., "Listening to Workers: Findings from the Commonwealth Fund 1999 National Survey of Workers' Health Insurance," The Commonwealth Fund, January 2000.

Feder, Judith, et. al., "The Difference Different Approaches Make: Comparing Proposals to Expand Health Insurance," The Kaiser Project on Incremental Health Reform, October 1999.

Fronstin, Paul, "Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 2000 Current Population Survey," Employee Benefit Research Institute Issue Brief 228, December 2000.

Glied, Sherry A. "Challenges and Options for Increasing the Number of Americans with Health Insurance," The Commonwealth Fund, December 2000.

Gruber, Jonathan, "Tax Subsidies for Health Insurance: Evaluating the Costs and Benefits," NBER Working Paper W7553, January 2000.

Kahn III, Charles N. and Ronald F. Pollack, "Building a Consensus for Expanding Health Coverage," Health Affairs, January/February 2001.

Kaiser Family Foundation and Health Research and Educational Trust, "Employer Health Benefits: 2000 Annual Survey."

Ku, Leighton and Matthew Broaddus, The Importance of Family-Based Insurance Expansions: New Research Finding about State Health Reforms, Center on Budget and Policy Priorities, September 5, 2000.

Lemieux, Jeff, David Kendall, and S. Robert Levine, "A Progressive Path Toward Universal Health Coverage," Progressive Policy Institute Policy Report, December 2000.

Long, Stephen H. and M. Susan Marquis, "Have Small-Group Insurance Purchasing Alliances Increased Coverage?" Health Affairs, January/February 2001.

Naifeh, Mary, "Dynamics of Economic Well-Being, Poverty 1993-94: Trap Door? Revolving Door? Or Both?" Current Population Reports P70-63, July 1998.

Subcommittee on Oversight of the Committee on Ways and Means, U.S. House of Representatives, "Report on Marketing Abuse and Administrative Problems Involving the Health Insurance Component of the Earned Income Tax Credit," June 1, 1993.

Thorpe, Kenneth E. and Curtis S. Florence, "Why are Workers Uninsured? Employer-Sponsored Health Insurance in 1997," Health Affairs, March/April 1999.

U.S. General Accounting Office, "Earned Income Tax Credit: Advance Payment Option Is Not Widely Known or Understood by the Public," GAO/GGD-92-26, February 1992.

U.S. General Accounting Office, "Private Health Insurance: Potential Tax Benefit of a Health Insurance Deduction Proposed in H.R. 2990," GAO/HEHS-00-104R, April 21, 2000.



Major Proposals:
Refundable Tax Credits for Individuals to Purchase Health Insurance

During the 106th Congress, several members introduced legislation to establish refundable tax credits for individuals to purchase health insurance. President Bush proposed a tax credit for health insurance during the campaign. (This tax credit was separate from the overall Bush tax cut package and is not considered part of that package). In addition, some policy and advocacy groups have developed various individual tax credit proposals. The chart below outlines the major legislative proposals introduced in the last session of Congress, as well as the plans developed by President Bush and the Progressive Policy Institute.(1) All these proposed tax credits are either fully or partially refundable, but they differ in other key areas. These areas are discussed below and presented in the chart on the pages that follow.

Maximum credit: Most of the tax credit proposals establish flat-dollar credits, with separate credit amounts for individual and family coverage. Some of the proposals index these credit amounts to the overall Consumer Price Index, while others do not. None of these proposals index the credit amount to any measure of medical or health insurance inflation. Because health care costs rise at a faster rate than prices in the economy in general, the value of the health insurance credit — and the level of health care coverage they would finance — would almost certainly erode over time. The one exception in this area is the proposal put forth by Rep. Jim McDermott, which provides a credit equal to 30 percent of the cost of the health insurance premiums; because it is expressed as a percent of insurance costs, the amount of the credit would rise automatically with increases in health insurance costs.

Eligibility: All of the credits would be available to individuals that do not have access to employer-sponsored health insurance. Some of the proposals also make the credit available to those that do have access to employer-sponsored insurance. Only the plans advanced by PPI and Rep. Norwood would allow a credit that could be used to pay for an employee's share of employer-sponsored insurance.

Income phase out: Many of the credits are available to individuals and families no matter what their income level. Four proposals limit use of the credit to people below a specified income level.

Advance payment option: Some of the proposals authorize advance payments of the credit during the year, either through the employer or directly to an insurer, in an attempt to address the mismatch between the time that health insurance premiums must be paid and the time that income taxes are filed. Plans that allow advance payments all appear to require reconciliation on the employee's tax return at the end of the year, with the IRS recapturing any overpayment of the credit.

Restrictions on health policies: Most proposals would allow credits to be used for any type of health insurance policy, other than long-term care insurance, dental, and other highly limited policies. These restrictions, however, are not the same as setting an acceptable minimum level of insurance that can be purchased with the credit. Only Rep. Stark's proposal includes standards of insurance.


Online Search of Non-Group Insurance Policies

To illustrate the insurance choices available in the non-group market, the Center on Budget and Policy Priorities conducted a selective online search of non-group insurance policies available in six cities (Butte, Montana; Austin, Texas; Cleveland, Ohio; Sacramento, California; New York, New York; and Jackson, Mississippi).(2) In four of the six cities, a variety of plans were available from several insurers, but consumers in Butte and New York City had very limited choices. The price of a standard comprehensive non-group policy, equivalent to a typical employer-sponsored insurance policy ranged from $4,200 in Sacramento to $9,800 in Butte.(3) The average cost — about $7,400 — is similar to the premiums for a "medium" family plan identified by the General Accounting Office and cited earlier in this paper.

We also examined policies that were available in the range of $2,000 to $2,500, which is about the amount available to families under the major tax credit proposals. In New York City, our online search could not identify any family policies available in that price range. In the other five cities, low-cost plans were available, but they were very limited insurance policies. These plans generally had high deductibles, high cost-sharing and narrow benefit coverage.

For example, the low-cost policy in Austin has a deductible of $2,500 per individual or $5,000 per family (if providers in the insurance plan's network are used; outside the network, the deductibles are twice as high). For a family with an annual income of $25,000, a deductible of $5,000 implies that it must spend 20 percent of its income before receiving any benefits from this plan. After that deductible is met, all expenses are covered by the plan, unless providers outside the network are used. In that case, there is 50 percent cost sharing until out-of-pocket expenses reach $10,000 per person or $20,000 per family. Prescription drugs, one of the most rapidly rising components of medical expenses, are not covered at all in this plan, nor are maternity expenses. Low-cost plans available in the other cities were also quite limited — in two cities, physician office visits were not covered at all.

The non-group insurance policies that were available in the $2,000 to $2,500 price range would do little to cover a family's routine medical needs (such as treatment for pneumonia or inner ear infections), preventive care (like prenatal care or immunizations) or medications (like antibiotics, insulin, or oral contraceptives). The low-cost policies were primarily useful for reducing the financial risks associated with high-cost hospital stays, but would still leave a moderate-income family burdened with substantial medical costs. Low-cost insurance plans, when available, are not comparable to the breadth of coverage that most Americans have come to expect through employer-sponsored insurance or Medicaid.

Table 3
Examples of Prices for Non-group Health Insurance for a Family in Six Cities
  Butte, MT Austin, TX Cleveland, OH Sacramento, CA New York, NY Jackson, MS
Annual price of a standard family plan(4) $9,780 $7,296 $7,944 $4,188 $7,224 $8,052
Price range of available family plans $4,848 - $9,780 $1,248 - $10,008 $1,040 - $10,188 $864 - $8,240 $7,224 $1,317 - $8,748
What type of plan could a family purchase for $2,000 to $2,500 per year? Primarily covers inpatient hospital expenses. Office visits, maternity and prescription drugs not covered. Deductible of $500 per person. 50% co-payment by family, after meeting deductible.(5) Deductible of $2,500 per person or $5,000 per family per year when using PPO network; $5,000/ $10,000 outside network. After meeting deductible, expenses covered 100% if using network; if outside network, then 50% co-payments with out-of-pocket expenses capped at $10,000 per person or $20,000 per family. Maternity and prescription drugs not covered. Primarily covers inpatient hospital expenses. Office visits, maternity care, prescription drugs, emergency room are not covered. Deductible of $2,000 per family, then 20% family co-payment of next $10,000. Deductible of $1,500 per member, with a limit of $3,000 per family. Deductible waived for office visits. 25% co-payment after meeting deductible and for all office visits. Additional $1,000 deductible for maternity care, then 25% co-payment. Prescription drugs covered with a co-payment of $10 for generic and $25 for brand-name drugs. No plans found in this price range.(6) Deductible of $2,500 per person or $5,000 per family per year. Office visits, prescription drugs and emergency visits covered 100% after deductible. Maternity coverage not included.
NOTE: All prices were for a family of four (a 43 year old man, a 40 year old woman, a 16 year old girl and a 10 year old boy, all nonsmokers), quoted by ehealthinsurance.com (except where noted). Prices may vary by age, gender, location of individuals and other factors. People with certain health conditions or practices (e.g., smokers) may be ineligible or require higher prices. Policies typically exclude coverage of pre-existing medical conditions.

End Notes:

1. The proposal developed by Families USA and the Health Insurance Association of America is not included here. It does not include an individual health tax credit. The proposal expands public health insurance programs and established a new employer credit. Charles N. Kahn III and Ronald F. Pollack, "Building a Consensus for Expanding Health Insurance," Health Affairs, January/February 2001.

2. We used insurance quotes from ehealthinsurance.com, a private health insurance broker. In some cases, we used data from quotesmith.com. We estimated prices for a standard family of four, all nonsmokers. Actual prices available for a family would vary with the age, gender, location and health status of the family members.

3. We selected comprehensive plans with a $500 deductible, with 20 percent co-payments for costs after the deductible had been met. This was a typical employer plan, based on data in Henry J. Kaiser Family Foundation and Health Research and Educational Trust, Employer Health Benefits: 2000 Annual Survey.

4. The "standard family plan" approximated typical coverage available from employer-sponsored insurance: $500 deductible and 80% of costs paid by the insurer and 20% paid by the family. The actual terms of policies varied and we selected plans that were closest to these criteria. When multiple plans of this general type were available, the price of the least expensive plan was displayed.

5. No plans in this price range could be found at www.ehealthinsurance.com. This plan listed by www.quotesmith.com.

6. No plans in this price range could be found at either www.ehealthinsurance.com or www.quotesmith.com.