^{Posted on September 16, 2008}

Cost And Coverage Implications Of The McCain Plan To Restructure Health Insurance

Achieving John McCain’s vision would radically transform the U.S. health insurance landscape, with negligible gains in numbers of covered Americans, and an expected decline within five years.

by Thomas Buchmueller, Sherry A. Glied, Anne Royalty, and Katherine Swartz
Health Affairs
Published online 16 September 2008

ABSTRACT:

Senator John McCain’s (R-AZ) health plan would eliminate the current tax exclusion of employer payments for health coverage, replace the exclusion with a refundable tax credit for those who purchase coverage, and encourage Americans to move to a national market for nongroup insurance. Middle-range estimates suggest that initially this change will have little impact on the number of uninsured people, although within five years this number will likely grow as the value of the tax credit falls relative to rising health care costs. Moving toward a relatively unregulated nongroup market will tend to raise costs, reduce the generosity of benefits, and leave people with fewer consumer protections. [Health Affairs 27, no. 6 (2008): w472-w481 (published online 16 September 2008; 10.1377/ hlthaff.27.6.w472)]

In public opinion polling, both democrats and Republicans rate health care as a key domestic issue in the upcoming election. Health care policy was an important issue in the Democratic primary races, and many voters are familiar with the outlines of the plan proposed by Sen. Barack Obama (D-IL). Less attention has been paid to the health care proposal of John McCain (R-AZ), the Republican nominee. In this paper we describe the likely impacts of the McCain plan on the level and stability of insurance coverage and on the health care costs faced by families.

Senator McCain’s health plan has three central features: withdrawing the current tax exclusion of employer payments for employer-sponsored coverage (in other words, taxing premiums paid by employers), introducing a refundable individual health insurance tax credit, and deregulating nongroup insurance by permitting the purchase of policies across state lines. We focus on the effects of these features here. Senator McCain’s plan also contains other elements, including the promotion of disease management, information technology (IT), medical liability reform, and pricing transparency, but details on these elements have not been released, and space limitations preclude us from considering them here.

A description of Senator McCain’s plan is available on his Web site and in several speeches he has given.1 These sources reveal a vision of what, from Senator McCain’s perspective, a desirable health care system would look like. At its heart, the system he envisions is one in which many more—perhaps most—insured Americans would buy health insurance and health services in a national, relatively unregulated, competitive market, either on their own or as members of fluid, voluntary associations, such as churches or clubs. Because this would be a radical departure from the current system, its likely effects deserve close attention.

Replacing The Tax Exclusion With A Refundable Tax Credit: Impact On Overall Rates Of Coverage

The tax exclusion for employer-sponsored health insurance provides a subsidy to people who obtain coverage through their jobs. A sizable economics literature shows that the tax subsidy greatly increases employers’ offers of health benefits.2 Some of this research indicates that employers’ decisions are most responsive to the “tax price” faced by their more highly compensated workers.3 However, to qualify for the tax subsidy, employers must abide by IRS nondiscrimination rules, which require firms to provide similar benefits to high- and low-wage employees. These rules have the effect of increasing the health insurance coverage of less skilled workers who work in firms that also employ highly skilled workers.4 The tax exclusion also strengthens risk pooling by creating an incentive for younger, healthier people to remain in employer-sponsored groups, where they effectively subsidize higher-risk workers.5 For these reasons, the tax exclusion has been described as “part of the glue that holds employment groups together as risk pools for purchasing health benefits.”6

Eliminating the tax exclusion would greatly reduce the number of people who obtain health insurance through their employers.7 This decline would be driven by three factors: the effective price of employer-sponsored coverage would increase, the nondiscrimination rules would no longer apply, and low-risk employees would have less incentive to remain in employer-sponsored groups.

Losses because of effective price increase for employer-sponsored coverage. Elasticity estimates from published studies can be used to predict the coverage reduction caused by the effective price increase. Anne Royalty provides an estimate that is typical of those found by researchers in this area: her results suggest that eliminating the income tax preference for health insurance would result in a 17 percent decrease in the share of workers who are offered health insurance by their employers.8 This translates to a net decrease of twenty-eight million Americans (one out of every six people with employer-based coverage) covered by employer-provided health insurance. Jonathan Gruber and Michael Lettau find a somewhat smaller effect, in part because they assume that large firms will not respond to the policy; however, their results still suggest that ten to sixteen million Americans would lose their employment-based coverage.9 Other analyses imply larger decreases in coverage for the self-employed and those in small firms.10

For our analysis, we took a middle-range estimate from these studies and assumed that the elimination of the income tax preference for employer-sponsored insurance would cause twenty million Americans to lose such coverage. We note, however, that the effect could be much larger. Studies suggest that many employers would be quick to drop health benefits in response to a major policy change, such as the McCain plan, that greatly altered the business case for offering benefits.11 Also, as we note above, these estimates account only for the price effect of eliminating the tax preference; they do not account for the number of low-wage workers who might lose employer-sponsored insurance when employers are no longer bound by the nondiscrimination rules, nor do they capture the impact of breaking up existing risk pools.

Gains because of take-up of nongroup policies. The net effect of the McCain plan on how many people have health insurance will depend also on two other outcomes: the number of people who leave the employer-sponsored system but then purchase nongroup coverage, and the nongroup coverage take-up decisions of currently uninsured people who would have access to the McCain tax credit. Again, elasticity estimates from economic studies can be used to model this take-up behavior.12 However, a difficulty with estimating net coverage effects is that health insurance available in the group and nongroup markets is not directly comparable. Administrative costs are much higher in the nongroup market; as a result, coverage currently purchased there is less generous on average than in the group market. Coverage estimates are sensitive to assumptions made about the cost and comprehensiveness of the typical policies that would be found in the nongroup market. More important, consideration of less generous coverage raises questions about what it means to be insured.13

Consider two entirely different approaches to modeling the coverage effects. One assumes that plans available in the nongroup market would be as generous as those currently observed in the group market. The higher cost of coverage obtained in the nongroup market, however, would cause some people not to purchase it, even though the new tax credit would partially offset the higher nongroup premium. At the other extreme, we could assume that the McCain plan would lead insurers to develop “bare-bones” insurance policies costing little more than the $5,000 tax credit for families (and $2,500 for individuals). If families considered price alone, some would choose to take up this coverage. However, for most uninsured families the benefits of such policies in terms of protection from financial risk and access to medical care would likely be very small, and take-up would be much lower than if plans were more generous.14

We take a middle road between these two extremes and assume that nongroup policies would both be less generous and have premiums that are one-third less than those of current employer-based plans. We estimate that with the tax changes proposed by Senator McCain, about twenty-one million people (including some people who are currently uninsured and some who would lose employer-group coverage) would take up nongroup coverage.15

Net effects and caveats. Weighing this increase in nongroup coverage against the twenty million people we assume would lose employer-based coverage results in an increase in insurance coverage of one million people. Because of the tremendous uncertainty in the estimates of employer and family behavior, we view this analysis as suggesting that initially there would be no real change in the number of people covered as a result of the McCain plan. However, people are likely to have far less generous policies than those they have today.

These net changes are roughly consistent with prior estimates of the impact of the Bush administration’s health plan—the model for the McCain plan. Estimates of the net change in coverage for the Bush plan range from three to nine million people newly insured.16 The difference between the estimates that show a modest increase in coverage and those, like ours, that show essentially no change reflects the fundamental uncertainty about how employers would respond to the enormous change in incentives associated with elimination of the income tax exclusion for insurance coverage. Models predicting increases in coverage generally assume that very few employers would drop health benefits.17 This assumption runs counter to most empirical evidence and to economic theory. In addition, some studies suggest much lower nongroup take-up, while others find that even sizable subsidies would not induce workers who have declined employer coverage to take up the offered coverage.18 Thus, the take-up of nongroup insurance could be much lower than we estimate, and the number of uninsured people correspondingly higher.

What is clear from these estimates is that the McCain plan will not enable many more Americans to obtain health insurance—and it certainly will not achieve universal coverage. By our calculations, upward of forty million Americans would be uninsured—and that number would likely grow over time. The estimates described above focus on the initial impact of the plan. Over time, a refundable tax credit would not automatically adjust as health care costs increase—which is quite different from the current tax exclusion of employer premium payments. Thus, the effectiveness of the tax credit in inducing people to buy coverage would inevitably decline over time. Even if the tax credit were indexed to the Consumer Price Index (CPI), if the annual growth in premiums exceeded CPI-measured inflation by 6 percent—as was the case between 1999 and 2007—the value of the credit would be eroded so much that in just five years, five million more people would be uninsured.

Moving Families Into The Nongroup Insurance Market

Despite initially having only a small effect on the number of uninsured Americans,the McCain health plan would have a strong effect on the private health insurance system. A key aspect of Senator McCain’s plan is to move people from employment-based insurance coverage to the nongroup market. This is the market where individuals and families who do not have employer- or government-sponsored coverage buy their insurance today.

At first glance, average premiums in the nongroup market often appear to be lower than premiums for group coverage. But the apparent advantage of nongroup coverage is an illusion. The reality is that providing coverage through nongroup plans is much more costly than providing that coverage through groups. Administrative expenses are twice as high in nongroup markets as in group markets.19 The costs are higher because insurers in this market spend considerable resources on medical underwriting, and economies of scale are lost. It is much more expensive to sell insurance to millions of individuals one individual at a time than it is to sell to a much smaller number of employer groups, each comprising thousands of employees. For a typical family that moves from group to individual coverage, therefore, the move to nongroup insurance will raise premiums for an identical policy by more than $2,000 per year. Shifting people into the nongroup market would not save money for most Americans. Rather, it would lead to increased spending on administrative costs and a decrease in the portion of health spending that actually goes to providing care.

One reason that nongroup plans appear less costly is that they offer less coverage. The typical deductible in nongroup plans is about $2,750, compared to about $1,000 for group policies.20 Coinsurance rates average 26 percent in nongroup plans, compared to 20 percent in a typical employer-based plan. For plans with copayments, the average copayment in the nongroup market is between $30 and $40 per doctor visit, well above that of group plans. Many services are not covered at all. Thus, much of the apparent savings from shifting to nongroup coverage would be offset by higher out-of-pocket costs for care.

A second reason that nongroup premiums appear low is that quoted prices are for coverage sold to healthier people. Insurers lose money when people are sick, so insurers in the nongroup market make every effort to avoid such people. A recent survey looks at the experience of people who are less healthy in nongroup markets.21 One-third of such people buying or looking into nongroup coverage were denied coverage or charged more because of a pre-existing condition. Nearly half found it difficult or impossible to find the coverage they needed, and more than two-thirds found it difficult or impossible to find affordable coverage. Those with chronic conditions are acutely affected by medical underwriting, although the practice is not limited to consumers most people would consider “sick.”22

Senator McCain has acknowledged the deep flaws in today’s nongroup market. He proposes two changes to make this market work better: state-based high-risk pools for the sick (called Guaranteed Access Plans), and the development of “innovative multi-year health insurance plans.” High-risk pools can be helpful as a way to cover some of the very sick. But to make more than a small dent in reducing the number of uninsured people, high-risk pools would need to be well funded. Senator McCain has proposed spending $7-$10 billion to subsidize high-risk pools—not nearly enough to make them work to cover significant numbers of the uninsured. No more than three million people would have affordable coverage with his proposed funding levels.23

State high-risk pools under Senator McCain’s plan are likely to go the way that current high-risk pools have gone. Currently, thirty-four states sponsor such pools, but they cover fewer than 200,000Americans.24 Because money is scarce, states deliberately restrict benefits and keep enrollment low. For example, Florida’s high-risk pool has been closed to new enrollees since 1991. California has rationed access to its pool by limiting the amount of time enrollees can be covered through the program and by capping enrollment. As a result, a sizable number of consumers who are too healthy to qualify for coverage through a high-risk pool will be deemed “high risk” by nongroup insurers and will face very high premiums, restricted benefits, and, in some cases, outright denials of coverage. Senator McCain’s plan to eliminate many state regulations that currently allow many high-risk people to participate in regular insurance plans would place an even greater burden on these fragile high-risk pools.

Senator McCain’s proposed multiyear health insurance plans also are unlikely to fix the shortcomings of the nongroup market. Indeed, these new plans are even unlikely to exist. True multiyear health insurance contracts, which promise people coverage for many years at a predetermined price, have never existed, and nothing Senator McCain has proposed would lead them to appear. The closest thing to a multiyear contract that does exist is a “guaranteed-renewable” policy. Under state insurance regulations that currently exist in all but three states (regulations that would be effectively eliminated under the McCain plan), individual insurance purchasers are guaranteed the ability to renew the same policy at a rate that applies to their entire rate class.25

Multiyear plans also have flaws. People do not want to be locked into the same health insurance plan year after year. When new medical services are developed, people want access to those services; an insurance policy must change to allow this. If people move, they want to be covered by new providers, not the providers in the town they moved from. Under guaranteed-renewable policies, only those who remain healthy can hope to switch coverage. For everyone else, a lock-in policy holds little value in a dynamic economy.

Currently, the nongroup insurance market is regulated at the state level, and local insurers, often Blue Cross and Blue Shield plans, are major players. Senator McCain’s proposal envisions a relatively unregulated national market for nongroup insurance in which families would buy insurance on their own or as members of fluid, voluntary associations, such as churches or clubs. In this market, consumers could select insurers licensed in any state. With more choice and competition, he believes that costs would fall and service quality would increase.

Everything we know about nongroup insurance markets, however, suggests that this vision is wrong. Health care is the ultimate local good: it is provided face to face, between doctor and patient. Today, most health plans negotiate contracts with local providers, directly or through intermediaries. The only truly national plans are traditional indemnity plans that do not negotiate with local provider networks. Such plans were once the backbone of American health care. They lost out to more tailored plans, however, because they could not compete effectively. Without an informed local network, their prices were higher and quality was lower. There is no reason to think that this has changed.

The main effect of establishing a national market would be to undo state laws designed to establish minimum levels of coverage and protect consumers. In a national market where state licenses are not required, insurers will charter in places where regulations are scarce—much like credit card companies do today. As a result, people guaranteed basic benefits today would find those benefits eliminated under the McCain plan. People in most states would lose access to procedural protections, such as requirements that disputed decisions by managed care plans be subject to external review.26 People also would lose access to many benefit protections. For example, forty-seven states now require mental health parity, forty-nine states require coverage of breast cancer reconstructive surgery, and twenty-nine require coverage of cervical cancer screening.27 All of these requirements—as well as regulations in several states that limit the rates that can be charged to higher-cost consumers and that limit who can be excluded from a health plan—would be eliminated under the McCain plan. Without legal requirements in place, plans would no longer offer these benefits at all in many markets, even if many consumers want them.28

Costs Of The Plan

It is difficult to estimate the cost to the federal government of Senator McCain’s health plan, since critical details of his plan have not yet been provided. The Brookings Institution and Urban Institute’s Tax Policy Center estimate that the tax-related provisions in the McCain plan would cost about $1.3 trillion over ten years starting in 2009.29 In addition, the Guaranteed Access Plans, or high-risk pools, envisioned in the plan would cost about $70-$100 billion over this period. Current estimates of the costs of the plan have focused only on government costs, but the plan also would lead to shifts in spending within the private sector. The McCain plan would shift coverage toward the nongroup market, lead to reductions in the comprehensiveness of coverage in that market through deregulation, and encourage employer-based coverage to become less generous as well. These changes would have the effect of shifting costs from insurance premiums toward out-of-pocket payments, and people with chronic or acute illnesses would likely incur much higher out-of-pocket health care costs than they do now.

Improving The Plan

Several modifications of the McCain proposal could preserve the basic structure of the plan while ameliorating some of its likely negative effects. First, implementing a publicly funded reinsurance program for the nongroup market, as Senator Obama has suggested in his plan, or a system of risk adjustments of that market, as had been suggested in the health reform plan of President George H.W. Bush in 1992, would make it much easier for higher-risk people to purchase nongroup coverage. These public risk-spreading mechanisms also would improve the stability of coverage in the nongroup market. Second, developing a set of health insurance purchasing organizations (which in many plans are variously titled networks, alliances, coalitions, and connectors) would enable people to more easily shop for and enroll in coverage in the nongroup market using their tax credits. Third, modifying the tax credits so that the value of the credit is greatest for those with low incomes would lead to an increase in the number of people who could afford to purchase coverage under the McCain plan. Finally, indexing the size of the tax credit to the cost of health care would check the extent to which coverage under the plan would erode over time.

Achieving Senator McCain’s vision would radically transform the U.S. health insurance system. His plan would alter the nature, source, and financing of coverage for the nearly 160 million Americans who now receive health insurance through their employers. We estimate that twenty million Americans—about one in every eight people with job-based coverage—would lose their current coverage as a result of the change in the tax treatment of coverage. Initially, this loss of job-based coverage would be offset by an increase in coverage in the nongroup market (although not necessarily for the same individuals). Within five years, however, the net effect of the plan is expected to be a net reduction in coverage relative to what would have been observed if the tax treatment of employer-sponsored coverage remains as it is now. The decline of job-based coverage would force millions of Americans into the weakest segment of the private insurance system—the nongroup market—where cost sharing is high and covered services are limited. Senator McCain’s proposal to deregulate this market would mean that people in it would lose protections they now have. These changes would diminish the security of coverage for most Americans, especially those who are not—or someday will not be—in perfect health.

NOTES

1. See McCain-Palin, “A Call to Action: McCain Health Care Plan,” http://www.johnmccain.com/healthcare (accessed 8 September 2008).

2. Some of the early work on this topic is reviewed in D. Cutler, “Public Policy for Health Care,” in Handbook of Public Economics, vol. 4, ed. A. Auerbach and M. Feldstein (Amsterdam: Elsevier, 2002), 2143-2243.

3. J. Gruber and M. Lettau, “How Elastic Is The Firm’s Demand for Health Insurance?” Journal of Public Economics 88, nos. 7-8 (2004): 1273-1293.

4. W.J. Carrington et al., “Nondiscrimination Rules and the Distribution of Fringe Benefits,” Journal of Labor Economics 20, no. 2 (2002): S5-S33.

5. A.C. Monheit, L.M. Nichols, and T.M. Selden, “How Are Net Health Insurance Benefits Distributed in the Employment-Related Insurance Market?” Inquiry 32, no. 4 (1995-96): 379-391; and T.M. Selden and D.M. Bernard, “Tax Incidence and Net Benefits in the Market for Employment-Related Health Insurance: Sensitivity of Estimates to the Incidence of Employer Costs,” International Journal of Health Care Finance and Economics 4, no. 2 (2004): 167-192.

6. A.C. Enthoven and S.J. Singer, “Market-Based Reform: What to Regulate and by Whom?” Health Affairs 14, no. 1 (1995): 105-119.

7. Altering the tax subsidy for insurance also has distributional consequences. Other things equal, replacing the current tax exemption for employer-sponsored insurance with a refundable credit increases the progressivity of the subsidy. However, as we discuss above, eliminating the nondiscrimination rules would likely have a negative effect on low-wage workers who currently receive health benefits through their employers.

8. A. Royalty, “Tax Preferences for Fringe Benefits and Workers’ Eligibility for Employer Health Insurance,” Journal of Public Economics 75, no. 2 (2000): 209-227.

9. Gruber and Lettau, “How Elastic Is the Firm’s Demand for Health Insurance?”

10. J. Gruber and J. Poterba, “Fundamental Tax Reform and Employer-Provided Health Insurance,” in Economic Effects of Fundamental Tax Reform, ed. H.J. Aaron and W.G. Gale (Washington: Brookings Institution, 1996), 125-170; and R. Feldman et al., “The Effect of Premiums on the Small Firms’ Decision to Offer Health Insurance,” Journal of Human Resources 32, no. 4 (1997): 635-658.

11. P. Fronstin, “The Future of Employment-Based Health Benefits: Have Employers Reached a Tipping Point?” EBRI Issue Brief no. 312 (Washington: Employee Benefit Research Institute, December 2007).

12. J. Gruber, “Covering the Uninsured in the U.S.” NBER Working Paper no. 13758 (Cambridge, Mass.: National Bureau of Economic Research, January 2008).

13. Prior experience with health insurance tax credits suggests that some insurers may seek to develop and market products of very little true value that sell for close to the amount of the tax credit.See House Ways and Means Committee, Subcommittee on Oversight, “Report on Marketing Abuse and Administrative Problems Involving the Health Insurance Component of the Earned Income Tax Credit,” WMCP: 103-14, 103rd Cong., 1st sess., 1 June 1993.We assume in our modeling that regulations implemented to avoid such fraudulent behavior would mean that insurance purchasers who used the credit would be required to make meaningful contributions toward the cost of their coverage.

14. S. Glied, “Is Something Better than Nothing? Health Insurance Expansions and the Content of Coverage,” in Frontiers in Health Policy Research, vol. 6, ed. D. Cutler and A. Garber (Cambridge, Mass.: MIT Press, 2003), 55-86.

15. We use Gruber’s preferred estimate of an average take-up elasticity of -0.5 in the nongroup market. The base is the current number of uninsured people (forty-seven million) plus the twenty million estimated to lose group coverage. See Gruber, “Covering the Uninsured in the U.S.”; and Gruber and Lettau, “How Elastic Is the Firm’s Demand for Health Insurance?”

16. R. Carroll, “The Economic Effects of the President’s Proposal for a Standard Deduction for Health Insurance,” National Tax Journal 60, no. 3 (2007): 419-431; Congressional Budget Office, “Appendix C: The President’s Proposal for a Standard Tax Deduction for Health Insurance,” in An Analysis of the President’s Budgetary Proposals for Fiscal Year 2008 (Washington: CBO, March 2007); and J. Sheils and R. Haught, “President Bush’s Health Care Tax Proposal: Coverage, Cost and Distributional Impacts” (Fairfax, Va.: Lewin Group, January 2007).

17. For example, the CBO and the Lewin Group estimate that “only” six to twelve million people would lose employer-based insurance under the Bush administration proposal. Ibid.

18. M.S. Marquis and S. Long, “Worker Demand for Health Insurance in the Non-Group Market,” Journal of Health Economics 14, no. 1 (1995): 47-63; and J. Gruber and E. Washington, “Subsidies to Employee Health Insurance Premiums and the Health Insurance Market,” Journal of Health Economics 24, no. 2 (2005): 253-276.

19. CBO, CBO’s Health Insurance Simulation Model (Washington: CBO, 2007).

20. Data for nongroup coverage are for 2006-07. Data are for preferred provider organization (PPO)/point-of-service (POS) plans, the most common option chosen. See America’s Health Insurance Plans, Individual Health Insurance 2006-2007: A Comprehensive Survey of Premiums, Availability, and Benefits (Washington: AHIP, 2007). Data for employer coverage are from Henry J. Kaiser Family Foundation/Health Research and Educational Trust, Employer Health Benefits: 2006 Annual Survey (Menlo Park, Calif.: Kaiser Family Foundation, 2006).

21. The sample is people ages 19-64 with individual coverage or who thought about it or tried to buy it in the past three years. See S. Collins et al., “Squeezed: Why Rising Exposure to Health Care Costs Threatens the Health and Financial Well-Being of American Families” (New York: Commonwealth Fund, September 2006). These findings are not unique. See K. Pollitz, R. Sorian, and K. Thomas, “How Accessible Is Individual Health Insurance for Consumers in Less-than-Perfect Health?” (Menlo Park, Calif.: Kaiser Family Foundation, 2001).

22. See, for example, D. Grady, “After Caesareans, Some See Higher Insurance Cost,” New York Times, 1 June 2008.

23. This estimate is based on the actual experience of the high-risk pools in California and Illinois, two large states with readily accessible data.

24. Health Insurance Resource Center, “Risk Pools for the Medically-Uninsurable,” http://healthinsurance.org/risk_pools (accessed 8 September 2008).

25. V. Patel and M.V. Pauly, “Guaranteed Renewability and the Problem of Risk Variation in Individual Health Insurance Markets,” Health Affairs 21 (2002): w280-w289 (published online 28 August 2002; 10.1377/ hlthaff.21.w280).

26. Information on external review is available from statehealthfacts.org, http://www.statehealthfacts.org, sponsored by the Kaiser Family Foundation, and is for 2008 (downloaded 22 August 2008).

27. V.C. Bunce and J.P. Wieske, “Health Insurance Mandates in the States 2008,” http://www.cahi.org/cahi_contents/resources/pdf/HealthInsuranceMandates2008.pdf (accessed 8 September 2008).

28. It is sometimes argued that state mandates are wasteful, forcing insurers to provide benefits that consumers don’t value,which raises premiums and lowers coverage. When adverse selection exists, however, unregulated competition among insurers can result in certain benefits’ not being available even though many consumers do value them. Furthermore, the most credible empirical studies on this issue provide no evidence that state insurance mandates have the effect of reducing coverage. See J. Gruber, “State Mandated Benefits and Employer Provided Insurance,” Journal of Public Economics 55, no. 3 (1994): 433-464; and J. Gruber, “The Incidence of Mandated Maternity Benefits,” American Economic Review 84, no. 3 (1994): 622-641.

29. L. Burman et al., An Updated Analysis of the 2008 Presidential Candidates’ Tax Plans: Revised August 15, 2008, http://www.taxpolicycenter.org/UploadedPDF/411749_update_candidates.pdf (accessed 27 August 2008).

Tom Buchmueller is the Waldo O. Hildebrand Professor of Risk Management and Insurance in the Ross School of Business, University of Michigan, in Ann Arbor. Sherry Glied (sag1@columbia.edu ) is a professor and chair of the Department of Health Policy and Management, Mailman School of Public Health, Columbia University, in New York City. Anne Royalty is an associate professor of economics, Indiana University—Purdue University at Indianapolis (IUPUI). Katherine Swartz is a professor of health economics and policy in the Department of Health Policy and Management, Harvard School of Public Health, in Boston, Massachusetts.