Available in Adobe Acrobat PDF format

The latest Census Bureau figures are in, and they show that the number of uninsured Americans increased last year. Although those estimates may be overstated, the problem remains serious: millions of Americans do not have health insurance. Senator John Kerry proposes a massive expansion of government programs and large subsidies for private insurance. This unprecedented commitment of taxpayer dollars would do little to slow the rapid rise of health care costs. In contrast, President George W. Bush's proposal is centered on tax credits for people to purchase their own coverage. By giving individuals ownership of their health insurance, the Bush proposal begins to change the incentives toward prudent purchasing and better value for our health care dollars.

The number of Americans without health insurance has moved upward once again. According to the latest survey from the U.S. Census Bureau, 45 million people were uninsured in 2003, an increase of almost 1.4 million from the year before.[1] About 15.6 percent of the population, or nearly one person in six, did not have health insurance last year. That is the highest rate of non-coverage since 1998, the last time the uninsured numbers spiked, when 16.3 percent were uninsured.

The news is disappointing, but not unexpected. The loss of health insurance reported for 2003 is the lingering effect of the 2001 recession. Although the economy was expanding, non-farm employment fell by 243,000 in 2003.[2] Since most people have employment-based health insurance, job losses often mean loss of insurance.[3] Other factors, including rising insurance premiums and a shift from full-time jobs to part-time or temporary jobs that do not offer health benefits, exacerbated the decline in coverage. The number of people with employment-based insurance fell by 1.3 million. Direct purchases of private insurance also decreased somewhat last year.

Both presidential candidates have advanced proposals to address the continuing problem of the uninsured. Senator Kerry proposes a big-ticket agenda of Medicaid expansions and subsidies to employers. If his figures are correct, that agenda might cut the number of people without insurance by 60 percent at a cost to taxpayers of some $650 billion--but a careful reading suggests that spending could exceed $1 trillion over ten years. President Bush proposes tax credits to help individuals purchase their own insurance coupled with insurance market reforms intended to make private insurance easier to obtain. The president's agenda represents a smaller commitment of taxpayer funds, perhaps $100 billion over the next ten years, and could lower the number of people without coverage by about 15 percent.

Would the candidates' health insurance proposals work as they predict? Has either candidate proposed sustainable reforms that would reduce health care costs and make insurance more affordable? After fifty years of debate in this country over health insurance that started with Harry Truman, have we at last come upon the solution to one of the most vexing policy problems of our time?

How Big Is the Problem?

Before adopting a major new policy initiative, it is important to have a clear idea of the dimensions of the problem we are trying to solve. The Census Bureau's estimate of the number of uninsured may seem precise, but it is an ambiguous indicator at best. That statistic is supposed to count people who are uninsured for the entire year, but many analysts believe the number more accurately reflects the uninsured on a given day.[4] Moreover, the uninsured population changes constantly as some people gain coverage while others lose it. One number cannot capture the varied experiences and circumstances facing people who do not have health insurance.

Piecing together information from a variety of Census Bureau and other sources, a picture of the uninsured emerges:

• Although most of the uninsured have low and moderate incomes, many live in high-income families and may be able to afford coverage. The Census Bureau reported that 33 percent of the uninsured were in households with incomes over $50,000 in 2003, and about half of those were in households with incomes over $75,000.[5] A more conservative estimate from Actuarial Research Corporation (ARC) showed that 20 percent of the uninsured were in families with incomes over $50,000 in 2002.[6] Note, however, that an uninsured person may not have access to the income attributed to the family or household.
• Some of the uninsured are eligible for government health coverage but do not enroll. Medicaid and the State Children's Health Insurance Program (SCHIP) are open to low-income people who meet other program requirements. Some people who are eligible do not enroll in those programs for a variety of reasons. Some perceive the stigma of a welfare program, others are daunted by the red tape, and others are unaware that they are eligible to enroll. ARC estimated that 3.8 million people, or about 9 percent of the uninsured in 2002, were eligible for Medicaid and SCHIP but did not enroll.[7]
• The Census Bureau underreports the number of people actually enrolled in Medicaid, which overstates the number of uninsured. ARC estimated that at least 10 million additional people were actually enrolled in Medicaid in 2002 but did not report this when surveyed.[8] That represents about 23 percent of the uninsured who were misclassified in the official statistics.
• Some of the uninsured are eligible for private insurance but turn it down. According to the Center for Studying Health System Change, 20 percent of all uninsured persons in 1997 were offered health insurance by an employer but did not enroll.[9] Those who decline employer coverage and remain uninsured tend to be low-income workers and their families.
• Young adults are less likely to have insurance than other age groups. The Census Bureau reported that 30 percent of people between eighteen and twenty-four years old were uninsured in 2003, compared with 18 percent of those between twenty-five and sixty-four. Most young adults are healthy and less likely to feel the need for health insurance, particularly if the cost of coverage is high relative to their incomes.
• Many of the uninsured are without coverage for fairly brief periods of time. The median duration of non-coverage for people uninsured at some time between 1996 and 1999 was 5.6 months according to a Census Bureau study.[10] The Congressional Budget Office (CBO) estimated that 45 percent of all spells without coverage in 1998 lasted four months or less, while 29 percent lasted a year or more.[11] Over the course of three years, 58 percent of people experiencing a spell of non-coverage had only one such spell.[12]

It is clear that the Census estimate of 45 million uninsured conceals as much as it reveals. Some 10 million people included in that count actually had insurance. Perhaps as many as 15 million others realistically could have obtained coverage through Medicaid, SCHIP, or private coverage through employers or the non-group market, and chose not to. Still, half of the uninsured--possibly 20 to 25 million people--do not have those options and are risking financial ruin and poorer health because they do not have health insurance.

The lack of health insurance is not a new problem that has suddenly emerged in the past few years. In 1987, the first year for which data are available, 12.9 percent of the population were uninsured. That number has grown steadily since then. Declines occurred only in 1999 and 2000, but even the lowest rate reported then--14.2 percent in 2000--is much higher than non-coverage rates observed in the 1980s.[13] Lack of health insurance is a serious matter for millions of Americans, and the political resolve to address the problem may be at hand.

Senator Kerry's Proposals

Senator John Kerry has advanced a package of government subsidies and expansions of government programs that he believes would cut health insurance costs and extend coverage to the uninsured.[14] Taking a lesson from previous reform efforts that failed to gain popular support, the Kerry agenda stays carefully within the framework of public and private health insurance as we know it today. Rather than forcing people to make changes in their insurance, the Kerry approach leaves that decision up to the individual or his employer. He has, however, introduced a new political concept: voluntary mandates on health insurance. Anyone wishing to take advantage of potentially large subsidies must do it Kerry's way.

Subsidize private insurance. Senator Kerry proposes a combination of subsidies and organizational changes to promote private health insurance. The proposal has two components:

• Reinsurance, under which the federal government would reimburse employer-sponsored plans for 75 percent of costs once a beneficiary has spent $30,000.
• Congressional Health Plan, a new subsidized purchasing arrangement modeled after the Federal Employees Health Benefits Program--in essence, FEHBP II.

The first component of the private insurance proposal, referred to as reinsurance, is simply a new subsidy to employers. It does not add significantly to the financial protection against risk already available to employers in the insurance market.

Health care costs for a small business can vary dramatically from year to year, possibly as the result of just one or two employees becoming seriously ill. That problem can be solved by pooling health costs across employers, which is what insurers do.

A small employer who purchases health insurance becomes part of a much larger insurance pool typically covering several thousand small businesses (and tens of thousands of employees and dependents). The aggregate costs of a large pool are very predictable. With medical costs for a family close to $8,000 a year, even a small insurer will have tens of millions of dollars in annual claims. Since a single individual with unexpectedly high costs is unlikely to have much impact on the overall financial results, most health insurers do not purchase reinsurance.

Employers who pay health care costs directly (and thus "self-insure") can buy private reinsurance, although some have no need for that protection. Larger employers are in much the same position as an insurance company because the aggregate cost of thousands of employees is quite predictable. Medium-sized employers who self-insure face greater uncertainty about cost and are more likely to buy private reinsurance.

The Kerry reinsurance proposal would subsidize the cost of employer-sponsored health benefits. However, the proposal would not significantly reduce the risk of sponsoring a health plan. Employers already manage that risk by purchasing insurance or private reinsurance or by being large enough to not require additional protection.

The new subsidy primarily will benefit people who already have insurance coverage, but some additional people will gain health insurance through this policy. Employers will be required to extend their health benefits to all employees. Many firms do not offer health benefits to part-time and temporary workers, but those firms would weigh the additional cost of benefits against the new subsidy. Some firms will decide to broaden the availability of their employee benefit. Others will decide that the cost of extending coverage exceeds the value of the new subsidy and decline the subsidy.

The Kerry proposal has made much of the requirement that firms pass the full amount of the subsidy through to the employees, which is expected to lower their insurance costs. That seems simple enough, but wait until the wave of lawsuits that will be generated by unavoidable ambiguities in interpreting the pass-through.

The new federal subsidy would blunt incentives on the part of employers and insurers to control wasteful spending for high-cost cases. Consequently, firms must be prepared to satisfy a government auditor demanding proof that an employee who spent more than $30,000 did so in a prudent and reasonable way--no overpayments or unnecessary use of health services. Without aggressive administrative oversight, health care spending and federal subsidy payments would grow more rapidly.

That could mean opening up the health records of individual workers and examining bills spanning the entire year, not just those incurred once the $30,000 threshold has been reached. Health plans and providers that participate in the Medicare program are very familiar with the burdens of verifying claims, and they have staff dedicated full time to keeping the administrative beast fed. It is neither easy nor inexpensive.

Once the bills are opened, the mischief begins. Why should the government pay a larger subsidy to an employer plan just because the plan agreed to higher fees? Why should the government pay a larger subsidy to a plan that does a poorer job of limiting unnecessary use of services? If the reinsurance scheme proved to be expensive, there would be calls from Capitol Hill to stop paying more to inefficient health plans (or plans that have higher costs, which is a faulty indicator of efficiency). That means tighter federal controls over the price of services and the way doctors treat patients, starting at the first dollar of health care spending.

The second component of the Kerry private insurance proposal is more promising. Employers who do not offer their own health benefits and individuals without employer coverage could purchase insurance through a new Congressional Health Plan modeled after the Federal Employees Health Benefits Program (FEHBP). Such an arrangement could give participants a wider range of health plans to choose from than most employers offer today. Wider choice means that individuals can select a plan that better matches their preferences for health care.

A "premium rebate" program would subsidize insurance under what could be called FEHBP II. An employer who participated would get a tax credit for contributions to employee premiums, and an individual who participated in FEHBP II on his own would also receive a tax credit if the premium exceeded six percent of his income.

The problem with FEHBP II is that it is not FEHBP. Health insurers who offer coverage to federal employees would be required to participate in FEHBP II, but the federal group would not be commingled with nonfederal participants. That separation was dictated by concerns that the new people seeking insurance under this plan would have substantially higher costs than federal employees, which could drive up their premiums in a single risk pool.

Even if costs per participant under FEHBP II are identical to costs under FEHBP, beneficiaries will face higher premiums under the new plan for equivalent benefits. FEHBP II subsidies offered under the Kerry proposal are substantially below the subsidy that federal employees receive--about 75 percent of the cost of premiums.

High premiums for a generous benefit package are likely to attract enrollees who use more health services than average. This "adverse selection" of participants would drive up premiums in subsequent years as healthier people drop out of FEHBP II. FEHBP has not experienced ill effects from adverse selection,[15] but the new program starts at a disadvantage since most people already have employer-sponsored coverage and will not join FEHBP II. The program could be stabilized, but that would require substantial increases in federal subsidies.

Expand government insurance programs. Under the Kerry plan, state Medicaid and SCHIP programs would enroll children in families under 300 percent of poverty ($56,550 for a family of four in 2004), parents with incomes under 200 percent of poverty ($37,700 for a family of four), and other adults below the poverty line ($12,490 for a couple). Such policies greatly expand the existing entitlement and impose new mandates on the states that could prove to be unworkable. The program expansion would also crowd out some private insurance, inducing some people to drop coverage they already have from their employers to take advantage of the heavily subsidized state insurance programs.

Such a major expansion of Medicaid would meet resistance from cash-strapped states without a new federal subsidy. To forestall such opposition, Kerry proposes a financial swap. The federal government, which currently pays about 57 percent of the cost of the Medicaid program, would pick up the full cost of all children enrolled in Medicaid. State governments would be relieved of their share of children's Medicaid cost, but they would have to pay their part of the cost of any additional adults who enrolled in Medicaid. States would be fully responsible for the cost of children and adults who are newly enrolled in SCHIP, which is financed by a federal block grant.

The Kerry swap is likely to be just the fiscal relief that every governor is hoping for. Between 2006 and 2014, the first years in which the Kerry plan could be effective, states will spend about $360 billion for children's health care in Medicaid, according to the CBO.[16] If the new policy were in place today, that money that would be paid by the federal government. The true figure could be larger, since more children would enroll in Medicaid under broadened eligibility standards.

States would have a strong incentive to enroll as many uninsured children as possible into Medicaid and little reason to restrain program spending since only the federal government would be liable. That would relieve financial pressure on local hospitals and other health care providers at no cost to the state. Although states would have to spend additional funds to cover adults who would become eligible for Medicaid and SCHIP, that extra spending would be modest compared to the financial benefit states would receive under the Kerry plan.

This program expansion is not such a good deal for the uninsured. Medicaid is a notoriously poor payer, and many physicians and other providers do not accept Medicaid patients or limit the number they will see. Many states have imposed cost-cutting measures that limit access to needed services. For example, at least fourteen states limit the quantity of drugs that can be dispensed by limiting the number of prescriptions a patient may have in a month or limiting the number of refills permitted.[17] Although there is usually an appeals process, such restrictions often pose a barrier to appropriate treatment for Medicaid patients.

Despite serious drawbacks, free Medicaid coverage would crowd out private coverage that some people otherwise would have had. Some workers with good family coverage through their employers would drop to a less-expensive policy that covers only the employee and rely on Medicaid to cover their children. Others would not enroll in their employer's plan at all, trusting that they can sign up for Medicaid when they have some medical bills to pay. In addition, some employers would be tempted to drop their health benefit if they think that many of their employees could take advantage of Medicaid under the new rules. Studies of previous Medicaid expansions suggest that 10 to 25 percent of people who enrolled would otherwise have been covered by private insurance.[18]

Impact of the Kerry health plan. Federal health spending under the Kerry plan would increase dramatically. One widely cited estimate suggests that the plan would cost $653 billion between 2006 and 2014, and give about 28 million uninsured people new coverage.[19] That estimate is overly optimistic. It includes savings proposals that are unlikely to be given full credit by the CBO, and several of the spending proposals could be substantially more costly than had been assumed.

A forthcoming AEI study will provide evidence that the impact of the Kerry health plan on the federal budget will likely exceed $1 trillion over the next decade. The plan also is likely to cover far fewer of the uninsured than is generally believed. Much of the new spending would be funneled to those who already have health insurance, and the Medicaid expansions will not be as popular with the uninsured as the earlier estimate assumes.

The Kerry health plan would reduce the amount many people pay for health insurance, but the plan would do little to change the underlying forces driving health costs (and health insurance premiums) skyward. Both the Medicaid expansion and the Kerry reinsurance subsidy would reduce incentives to limit unnecessary spending. Although the plan expands health insurance choices for consumers under the FEHBP II proposal, the clear policy preference is to bolster the role of employers and the government in deciding what health insurance most people would be allowed to purchase.

President Bush's Proposals

President Bush has advanced three proposals to help the uninsured obtain health coverage: a refundable tax credit, an above-the-line income tax deduction (that can be used even by people who do not itemize their deductions) for high deductible health plans, and association health plans (AHPs).[20] The tax credit and association health plans have been part of the Bush agenda over the course of his term in office. The tax deduction for high deductible plans is a new proposal that keys off recently enacted health savings accounts (HSAs).

A refundable tax credit would provide a subsidy directly to individuals who buy their own health insurance. The credit would cover up to 90 percent of the insurance premium. The credit would be capped at $3,000 for a family and $1,000 for an individual and would phase out for those with higher incomes. The policy provides a benefit to low-income people who do not have the tax advantage of employer-sponsored coverage. They could use the credit to purchase insurance in the non-group market or through private or state purchasing pools, which would give them the benefit of group purchasing. Because such individuals have little or no tax liability, the credit would be refundable and available at the time the insurance was purchased.

Under the new HSA rules, a person who buys a qualified health insurance plan with at least a $2,000 deductible for a family policy and $1,000 for a single policy is now allowed to contribute money on a tax-preferred basis to a health savings account. The premiums for that insurance must be paid out of pre-tax dollars, however. The Bush proposal would allow an above-the-line deduction for those premiums. That would provide a new incentive for people to buy high-deductible insurance and increase their savings through an HSA. People who buy health insurance on the non-group market do not receive the favorable treatment given to those who purchase employer-sponsored coverage under current law.[21] The Bush proposal would place both groups on more even footing.

The HSA premium deduction proposal would enhance consumer awareness of health costs through higher-deductible insurance. People who have low-deductible coverage and modest co-payments act as if someone else is paying for much of their care. Consequently, they have less of an incentive to scrutinize carefully whether they are getting good value, and they are likely to overuse health services. Not surprisingly, raising deductibles brings down the cost of premiums. As a result, 61 percent of all policies sold on the non-group market have deductibles greater than $1,000, and 34 percent have deductibles greater than $2,000.[22]

One of the criticisms of tax credits for individual insurance is that such non-group coverage is often more expensive than comparable insurance purchased in the group market. The Bush association health plan proposal attempts to ameliorate that problem by permitting businesses, associations, and individuals to join together to purchase health insurance at group rates. AHPs would be subject to federal regulation and would be exempt from state benefit mandates.

According to the U.S. Treasury, the refundable tax credit and deduction for high-deductible insurance premiums would reduce federal revenues by $97 billion between 2005 and 2014.[23] The AHP proposal would have minimal ten-year cost.[24] About 6 million people would be newly insured under the Bush plan.[25]

The Bush plan has been criticized for not doing enough for the uninsured, and that might be a fair criticism. Many of the uninsured have low incomes and would need substantial assistance to buy coverage. Compared with the $400 billion Medicare prescription-drug benefit, the amount proposed for the uninsured looks small. Perhaps Congress should have spent less for those who have health insurance and more for those do not.

What may be small in dollar terms can still be a big idea, however. The Bush approach begins to put the consumer in the driver's seat. Individual tax credits would help the uninsured buy health coverage of their own choosing. Incentives to invest in HSAs would promote personal saving to defray the cost of health care. Both policies directly involve the individual in how their health care dollars are spent. Under this plan, people would begin to feel a sense of ownership and personal control over health financing decisions, and they would begin to demand better value for their dollars.

Admittedly, the non-group insurance market remains difficult to navigate even in this era of eHealthinsurance.com and similar websites. State regulations limit the kind of health insurance that can be sold and impose requirements that add to the cost of coverage. Consumers need better information and better choices, and they need a convenient way to buy their health insurance. Both Senator Kerry, in his FEHBP II proposal, and President Bush, in his AHP proposal, recognize that insurance market reform is a necessary part of the policy mix.

We obviously have a long way to go to solve the problem of the uninsured. President Bush proposes a less extensive program to expand individual coverage, specifically targeting the uninsured. He offers a radical idea: let people who need health insurance decide for themselves what coverage to buy. In contrast, Senator Kerry would spend a great deal of money to expand the current system of employer-sponsored coverage and government health programs. Much of that money will go to people who already have good employer coverage. Given the fiscal problems facing the next president, Senator Kerry may be biting off more than the country can chew.

Notes

1. Carmen DeNavas-Walt, Bernadette D. Proctor, and Robert J. Mills, Income, Poverty, and Health Insurance Coverage in the United States: 2003, U.S. Census Bureau, Current Population Report P60-226, August 2004.

2. Rachel Krantz, Marisa Di Natale, and Thomas J. Krolick, "The U.S. Labor Market in 2003: Signs of Improvement by Year's End," Monthly Labor Review, March 2004: 3-29.

3. John Cawley and Kosali I. Simon, "Health Insurance Coverage and the Macroeconomy," Economic Research Initiative on the Uninsured/University of Michigan, ERIU Working Paper 24, October 2003; and "Rising Uninsured Rates: It's Employment, Not the Economy, Stupid," ERIU Research Highlight no. 5, April 2004.

4. The Census Bureau acknowledges that point; see DeNavas-Walt, et al., Income, Poverty, and Health Insurance, 14. The primary Census Bureau estimate of the uninsured derives from the Current Population Survey (CPS), which asks questions every March about insurance status during the preceding year. The CPS does not measure how long a spell of non-coverage is for a person reporting that he is uninsured. Bhandari's recent analysis suggests that the CPS measure may more closely reflect the insurance status of people who have been without insurance for eight months or more, but the author indicates that factors other than a failure to report short spells on non-coverage could be involved. See Shailesh Bhandari, People with Health Insurance: A Comparison of Estimates From Two Surveys, U.S. Census Bureau, SIPP Working Paper no. 243, June 8, 2004.

5. DeNavas-Walt, et al., Income, Poverty, and Health Insurance, Table 5.

6. Author's calculations based on ARC analysis reported in BlueCross BlueShield Association, Increasing Opportunities for the Uninsured, 2004: 5.

7. Author's calculations based on ARC analysis reported in BlueCross BlueShield Association, Increasing Opportunities, 7.

8. Ibid., 5.

9. Peter J. Cunningham, Elizabeth Schaefer, and Christopher Hogan, Who Declines Employer-Sponsored Health Insurance and Is Uninsured?, Center for Studying Health System Change, Issue Brief no. 22, October 1999: 1. This figure includes both the adult employee and any dependents.

10. Shailesh Bhandari and Robert Mills, Dynamics of Economic Well-Being: Health Insurance 1996-1999, U.S. Census Bureau, Current Population Report P70-92, August 2003, fig. 11.

11. Congressional Budget Office, How Many People Lack Health Insurance and For How Long?, May 2003, Table 3.

12. Ibid., Table 5.

13. It is not clear how much good news those rate declines in 1999 and 2000 actually represent. The Census Bureau reported five different estimates of the uninsured in those two years, reflecting major changes in survey methodology. Some of the declines might be the result of those methodological changes, rather than being purely an improvement in health insurance coverage.

14. See "John Kerry's Plan to Make Health Care Affordable to Every American," available at www.johnkerry.com/issues/ health_care/health_care.html.

15. Curtis S. Florence and Kenneth E. Thorpe, "How Does the Employer Contribution for the Federal Employees Health Benefits Program Influence Plan Selection?" Health Affairs, March/April 2003: 211-218.

16. Author's calculation based on Congressional Budget Office, "Fact Sheet for CBO's March 2004 Baseline: Medicaid and the State Children's Health Insurance Program," available at www.cbo.gov/factsheets/2004b/Medicaid.pdf.

17. Jeffrey S. Crowley, Deb Ashner, and Linda Elam, Medicaid Outpatient Prescription Drug Benefits: Findings from a National Survey, Kaiser Commission on Medicaid and the Uninsured, December 2003: iv.

18. Academy for Health Services Research and Health Policy, Understanding the Dynamics of "Crowd-Out," June 2001: 13-18; and Douglas Holtz-Eakin, The Uninsured and Rising Health Insurance Premiums, statement before the Health Subcommittee, Committee on Ways and Means, U.S. House of Representatives, March 9, 2004: 12.

19. Kenneth E. Thorpe, "Federal Costs and Savings Associated with Senator Kerry's Health Care Plan," August 2, 2004, available at www.sph.emory.edu/hpm/thorpe/kerry8-23-04.pdf.

20. "Making Health Care Coverage More Affordable," available at www.whitehouse.gov/infocus/healthcare/index.html; and "Helping the Uninsured," Budget of the United States Government, Fiscal Year 2005, U.S. Government Printing Office, 2004: 151-152.

21. Note that the self-employed may also claim full deductibility for health insurance under current law.

22. Thomas D. Musco and Thomas T. Wildsmith, "Individual Health Insurance: New Studies Shed Light on Issues of Affordability, Access, and Plan Design," Healthplan, January/February 2004.

23. U.S. Treasury, General Explanations of the Administration's Fiscal Year 2005 Revenue Proposals, February 2004: 21-26.

24. Congressional Budget Office, Cost Estimate: H.R. 660, Small Business Health Fairness Act of 2003, July 11, 2003, indicates that AHPs would have offsetting effects on revenue and Medicaid spending. The net effect would be to increase the federal deficit by about $320 million between 2004 and 2013.

25. Council of Economic Advisers, "Health Insurance Credits," February 14, 2002, available at www.whitehouse.gov/ cea/ HealthCredit_Feb02wp.pdf.  The figure refers to the impact of the tax credit for health insurance only.  The current Bush health plan adds the deduction for high-deductible insurance, but total funding for the health plan is not much larger than funding for the tax credit of two years ago.

Joseph Antos is the Wilson H. Taylor Scholar in Health Care and Retirement Policy at AEI and the coauthor of Private Discounts, Public Subsidies: How the Medicare Prescription Drug Discount Card Really Works.