Question for Ms. Grace-Marie Turner, Galen Institute, from the Honorable Adrian Smith, following the House Budget Committee Hearing on “The Tax Code and Health Insurance”, October 18, 2007.
As we look at ways to use the tax code to reduce the number of uninsured, we have seen different proposals including either tax credits or tax deductions to individuals and families for health care. In a general sense, as we look at the differences between credits, deductions, or a combination of the two, which approach will 1) be more fiscally responsible; and 2) do more to reduce the number of uninsured?
Response from Grace-Marie Turner
Thank you for your question, Mr. Smith. As I discussed in my testimony, many members of Congress from both sides of the aisle have offered proposals that would move public policy forward regarding the tax treatment of health insurance. Ranking Member Rep. Paul Ryan, for example, is developing a proposal that would provide a universal tax credit for health insurance. President Bush has offered a proposal to replace the current tax exclusion with a generous universal tax deduction. Others have offered proposals for income adjusted, refundable tax credits. And some are considering a combination of a tax deduction and credit. Senator Hillary Clinton in her recent health proposal recommends capping the amount of income that higher-income employees can exclude from taxes through health insurance. And Sen. Ron Wyden has received a great deal of attention for his proposal to replace the current tax exclusion for job-based health insurance with a direct, income-adjusted subsidy to individuals.
The Health Policy Consensus Group, a group of leading health policy experts from the market-oriented think tanks, has long advocated addressing the tax treatment of health insurance, and many of our members support refundable tax credits for health insurance.
President Bush’s proposal in 2007 to allow a universal tax deduction brought a new idea to the table in allowing a generous deduction for health insurance combined with a credit against payroll taxes. Because all workers pay payroll taxes, this latter proposal would provide help to those at the lower end of the income scale who may not owe income taxes or are in lower tax brackets.
There are always going to be constituencies that argue for credits over deductions and vice versa and a case can be made for both. I personally believe that a combination of credits and deductions would be most beneficial and believe that merging them into a single policy initiative might provide the impetus to finally move policy forward on this important issue.
The refundable credit would be more valuable to those at the lower end of the economic scale by providing meaningful help to purchase health insurance. And a deduction would be more like the tax benefit which those with job-based insurance currently receive through the tax exclusion. Alternatively, Congress could cap the tax exclusion for job-based health insurance in order to limit the open-ended tax benefit it provides to those with higher incomes and the most generous health benefits.
If you were to develop a new system of subsidies, the amount could flow from the numbers to determine the amount of the credits and deductions and what the cutoff and trigger points would be. It even may be possible to give people the option to choose between the two.
See below for a discussion in the academic literature about the issue of credits vs deductions for health insurance. I am greatly indebted to my colleague Thomas Miller of the American Enterprise Institute for providing me with the following excerpts and citations from the academic literature.
Thank you again for the opportunity to testify before your committee on this important issue and for this follow-up question. I would be very happy to come in to discuss this research and these options further with you. As you will see, there clearly is much more room for research on this issue.
Thomas Miller, “Expanding Access to Care by Empowering Workers with Better Incentives and New Options,” in Covering America: Real Remedies for the Uninsured, Washington, D.C.: Economic and Social Research Institute, November 2002, online at http://www.cato.org/research/articles/miller-coveringamerica.pdf.
The primary vehicle for accomplishing various market-strengthening reforms that lower future health care costs and expand access to health care would be a new federal tax credit option. The tax credit would amount to 30 percent of the cost of qualified insurance coverage…Essentially, individuals could subtract this portion of their insurance costs directly from their federal income tax liability. The tax credit is an option; it would not eliminate the current tax exclusion that is available for workers insured by employer-sponsored insurance (ESI) plans. (A similar federal income tax deduction also is available on a partial basis—70 percent of the cost of qualified health insurance—for the self-employed, and it will become 100-percent deductible from federal taxable income in 2003.) Instead, it would provide a competitive alternative to the tax exclusion for those workers to opt for in place of the tax exclusion. It would encourage a more gradual transition toward other forms of private insurance coverage. Workers who choose to enroll in an ESI group plan would continue to use the current tax exclusion. Employees who choose to decline ESI coverage and not take advantage of the current tax exclusion could use the tax credit option instead to purchase other forms of health insurance coverage.
From Mark V. Pauly and Bradley Herring, Cutting Taxes for Insuring: Options and Effects of Tax Credits for Health Insurance, Washington, D.C.: AEI Press, 2002, full text and other information available at http://www.aei.org/books/bookID.45,filter.all/book_detail.asp.
…The Key Tradeoff
Our simulation estimates serve to illustrate numerically a key tradeoff suggested earlier. For a given amount “spent” on credits, there is a tradeoff between the breadth of the reduction in the number of uninsured and the depth of the increase in the coverage they take. There is also an interaction with risk levels. At one extreme, a flat credit that does not specify a minimum policy will cause all of the previously uninsured to obtain some insurance coverage. At very low risk levels, the previously uninsured will probably be able to buy coverage society would regard as “adequate.”(There is no objective standard for “adequate coverage.”)
But persons with high risks who are unwilling or unable to pay more of the premium themselves will have to select coverage with deductibles and (especially) upper limits. While the new coverage will provide both more protection against out-of-pocket payments and more encouragement for the use of beneficial care, the protection and encouragement will obviously be smaller than if nominal coverage were more generous.
Under a policy of fixed-dollar credits and a requirement to buy an “adequate” benchmark policy, some of the uninsured will reject the subsidy and remain uninsured. Persons with lower risks and those who place high value on avoiding being a charity or bad debt case will move to coverage which, by definition, is “adequate.” Compared to the alternative policy discussed in the previous paragraph, this policy will convert fewer people from uninsured to insured, but among those who are converted we will see a larger effect on their use of and protection by health insurance.
Finally, a policy of proportional credits will move fewer people out of the ranks of the uninsured, but, of those it does cause to become insured, more will come from the higher risk categories. But such a policy may also stimulate (and subsidize) the purchase of coverage in excess of the benchmark level; it could lead to “lavish plans,” especially among those who were formerly insured but can become eligible for the credit.
Which of these three alternatives is best? The answer clearly cannot be given with objective certainty; it all depends on how the different patterns of changes are valued. If one invokes the principle that the first few dollars of insurance coverage (like the first few dollars of anything beneficial) are likely to do the most good, a design that places rather light obligations on the comprehensiveness of coverage and uses fixed-dollar credits may make sense. But ultimately the choice itself will require consensus on exactly why “we” want the uninsured to become insured, and what benefits we expect to accrue to all from that change.
Another key issue when choosing tax credit options is how generous the credit is to be. At a given income level, small credits will have little effect on the number of uninsured, whereas large credits will have large effects. If we focus on the large majority of the uninsured who have incomes above the poverty line, our general conclusion is that credits will need to be substantial to make much of a dent in the number of uninsured. For low-income workers (and their dependents) below 300 percent of the poverty line (where the uninsured are disproportionately found), we conclude that substantial reductions in the numbers of uninsured will require credits in the range of approximately half of the individual insurance premiums, with even greater credits needed for families with incomes at the bottom of this range. Thus another important tradeoff occurs between reductions in the number of the uninsured versus tax revenues that could be spent on other public programs.
But note that much of the “cost” of tax credits does not represent a reallocation of real resources away from other uses and toward the health care needs of the previously uninsured. Instead, much of the credit effectively represents a tax reduction for the majority of lower-middle-income people who formerly had obtained health insurance for themselves and their families in some fashion. Limiting eligibility for the credit to a subset of those at the same income level engaging in the same health insurance purchasing behavior can reduce the “cost,” but at the real expense of horizontal inequity and substantial distortion in the labor market.
To make any such judgments rationally, however, one would need more information than just a head count of the formerly uninsured. The missing piece of information is important for the entire policy exercise: How much of an improvement in health is generated by the presence of insurance coverage (compared to its absence) for people at different income and risk levels? It is possible, for example, that insurance coverage for people who are initially low risks may produce more of an improvement in health than coverage for those who are initially high risks. Almost all of the research on the impact of insurance coverage either looks at the uninsured as a group or singles out poor uninsured people, but the most relevant question is the amount of good that health insurance would produce for a lower- middle-income family (compared to their being uninsured). As noted elsewhere by Pauly and Reinhardt (1996), our failure as researchers to produce this information on effectiveness makes it more difficult to persuade our fellow citizens to support tax credits or any other programs to reduce the numbers of the uninsured.
The fiscal design of tax credit programs is not the only influence on the number of uninsured. Most programs envision making everyone who is uninsured (at some income level) eligible for subsidy. This design stands in strong contrast to the Medicaid program, for which only some low-income uninsured are eligible. The universal character of tax credit programs would thus allow the government to direct subsidies or credit vouchers to everyone below a certain income level who is not insured; it would not be necessary for people to apply. In addition, once people at some income level had all been made eligible for credits judged to provide adequate subsidies to permit them to afford insurance, there would be less justification for someone to remain uninsured, and therefore less need to have a permissive charity care or bad debt policy applied to that person. Changes in the financial responsibilities imposed on uninsured people might themselves stimulate people to become insured, although some safety net will need to remain for those who truly fall through the cracks. Finally, rewarding the great majority of lower-middle-income people who do choose to be insured with a substantial tax reduction might both call attention to the social value of being insured and offer the uninsured further incentive to change their status. While it is unlikely that the number of uninsured will ever be literally zero, carefully designed credit programs can both reduce the numbers of uninsured and improve the equity of tax treatment of the insured.
Footnote citation from Lily L. Batchelder, Fred T. Goldberg Jr., and Peter R. Orszag, “Efficiency and Tax Incentives: The Case for Refundable Tax Credits,” New York University School of Law, NYC Center for Law and Economics, November, 2006, at http://ssrn.com/abstract=941582.
For example, Zelinsky has discussed why tax incentives may enhance economic efficiency by correcting for positive externalities and has applied his analysis to the home mortgage interest deduction and accelerated depreciation. See Edward A. Zelinsky, Efficiency and Income Taxes: The Rehabilitation of Tax Incentives, 64 TEX. L. REV. 973 (1986). However, his article is concerned with the decision of whether to retain or institute a tax incentive and not with what form of tax incentive is most efficient. Id. at 1023. Weiss has proposed converting certain investment tax incentives to credits, but this proposal was based on equity and not efficiency concerns. See Deborah M. Weiss, Tax Incentives Without Inequity, 41 UCLA L. REV. 1949 (1994).
The economics literature has examined the merits of credits relative to deductions only in a few specific examples, such as the deduction for charitable giving and the exclusion for gifts. See, e.g., Louis Kaplow, A Note on Subsidizing Gifts, 58 J. PUB. ECON. 469 (1995) (considering the optimal subsidy for gifts in the presence of externalities and concluding that a tax deduction is not obviously inferior to a credit); Peter Diamond, Optimal Tax Treatment of Private Contributions for Public Goods With and Without Warm Glow Preferences, 90 J. PUB. ECON. 897 (2006) (concluding that the optimal subsidy for private contributions to public goods may rise with earnings but not reaching any policy conclusions about whether the optimal subsidy is a deduction). Rosen provides a brief general discussion of the choice between a deduction and a credit when the purpose of a provision is to encourage certain behavior. See HARVEY ROSEN, PUBLIC FINANCE 377 (7th ed. 2005) (“If the purpose is mainly to encourage certain behavior, it is unclear whether credits or deductions are superior . . . . If people differ with respect to their elasticities of demand, it may make sense to present them with different effective prices.”). Gruber also provides a brief general discussion of the choice between deductions and credits but does not discuss the possibility that externalities or elasticities vary by income level. See JONATHAN GRUBER, PUBLIC FINANCE AND PUBLIC POLICY 507 (2005). He also briefly discusses the debate about refundability but only from an equity and not an efficiency perspective. See id. at 508. Seidman argues for converting some deductions and exclusions into refundable credits but largely on equity grounds. LAURENCE S. SEIDMAN, POURING LIBERAL WINE INTO CONSERVATIVE BOTTLES 20-27 (2006). He also differs in his skepticism of uniform credits, arguing that “a better prescription would simply be [that] . . . each refundable tax credit should utilize a schedule that the citizenry judges to be equitable.” Id. at 26.
Additional literature citations that may be useful to you include:
James D. Reschovsky and Jack Hadley, “The Effect of Tax Credits For Nongroup Insurance on Health Spending By the Uninsured,” Health Affairs, February 25, 2004, at http://content.healthaffairs.org/cgi/content/full/hlthaff.w4.113v1/DC1.
Mark Pauly and Bradley Herring, “Cutting Taxes for Insuring: Options and Effects of Tax Credits for Health Insurance,” University of Pennsylvania, May, 2000, at http://council.brandeis.edu/pubs/Paulytx.PDF
Jonathan Gruber, “Coverage and Cost Impacts of the President's Health Insurance Tax Credit and Tax Deduction Proposals,” Kaiser Family Foundation, March 2004, at http://www.kff.org/insurance/upload/Coverage-and-Cost-Impacts-of-the-President-s-Health-Insurance-Tax-Credit-and-Tax-Deduction-Proposals.pdf.